CRO Buyer's Guide: Choosing a Chemistry Partner

This CRO buyer’s guide is for the person who has a molecule, a budget, and a date, and now has to choose a chemistry partner from a list of vendors that all describe themselves in almost identical language. Every website says custom synthesis. Every website lists HPLC and NMR. Every website mentions scale-up. None of that tells you which one can actually make your compound on time.

The useful distinctions are elsewhere. They are in whether the organization is structured to do research or to manufacture, whether the technique your route depends on is currently staffed rather than merely installed, whether the quality posture matches your actual regulatory stage rather than the most expensive stage available, and whether the person who answers your technical questions is the person who will run the chemistry.

This is the hub page for our CRO selection work. It sets the framework and links down to the detailed posts on pricing, RFQs, specific techniques, and the scale-up decisions that follow. Read this one first, then follow the links that match your situation.

What a Chemistry CRO Actually Is

A chemistry CRO is an organization that sells research and development labor and expertise, not manufactured product. You are buying problem-solving capacity: a route to a molecule, a method that resolves an impurity, a set of analogs for a structure-activity campaign. The deliverable is often material, but the value is the knowledge that produced it.

That definition matters because it sets what the organization is optimized for. A CRO is staffed with development chemists, runs many small projects in parallel, and is comfortable with a project whose specification changes at week three. Its equipment is flexible and its batch records are lab notebooks. It is not built to make the same intermediate every month for four years.

CRO versus CDMO versus CMO

Buyers conflate these three constantly, and it causes mis-scoped requests for quote more often than any other single error. The distinction is not about size or quality. It is about which part of the lifecycle the organization is built for.

	CRO	CDMO	CMO
Core offer	Research and development services	Development plus manufacturing	Manufacturing only
Typical deliverable	Route, method, data, small quantities	Validated process plus clinical or commercial supply	Product to your process
Regulatory posture	Usually non-GMP, GMP-capable in places	GMP throughout, filing support	GMP, executes your filed process
Scale range	Milligram to multi-kilogram	Kilogram to metric ton	Campaign scale, defined
Best fit stage	Discovery to preclinical	IND through commercial	Commercial, stable process
How they price	FTE, fee-for-service, milestone	Development fees plus per-kilogram supply	Per kilogram, campaign minimums

The practical test: if you do not yet know how to make the compound, you need a CRO. If you know how to make it and need it made under a filing, you need a CDMO. If you have a validated process, a filed specification, and simply need volume, you need a CMO.

Sending a CMO a discovery-stage request produces either a refusal or a quote priced for a validated campaign. Sending a CRO a commercial supply request produces a quote for a development program you did not want. Neither response is dishonest; both are the predictable result of an ambiguous scope. Our full treatment of that boundary is in CRO vs CDMO: a decision framework, and the honest version of where a mid-size US CRO wins and loses against a large CDMO is in ChemContract vs large CDMOs: an honest comparison.

The Hybrid Reality

Most real organizations sit somewhere between these categories, and the labels on websites are marketing rather than taxonomy. What matters is not the label but the answer to two questions: which of these three activities do you actually staff today, and which of them is the majority of your revenue. An organization that describes itself as a CDMO but earns most of its revenue from discovery-stage synthesis will behave like a CRO, which may be exactly what you want. Ask directly, and ask early, because the answer changes what you should send them.

Why the Distinction Changes Your RFQ

Each provider type reads a request for quote differently. A CRO reads a structure and asks how it would make it. A CMO reads a specification and asks what the annual volume is. If your request contains a structure but no volume, a CMO cannot quote it. If it contains a volume but no route, a CRO will quote development work you may not have budgeted for. Write the request for the type of organization you have decided you need.

The Five Buying Situations

Almost every chemistry outsourcing request falls into one of five situations, and each one needs a different provider profile, a different specification, and a different contract. Identify yours before you write the request for quote.

Situation 1: One-Off Compound for Screening

You need 100 milligrams to 5 grams of a compound that is not commercially available, for an assay or an early structure-activity series. Speed matters more than cost per gram. Purity of 95 percent by HPLC with confirming NMR is usually sufficient, and you rarely need a full impurity profile.

What to look for: a provider with a broad catalog and a stocked building-block inventory, because the fastest path is often two steps from something already on the shelf. Check whether the compound or a close precursor already exists in a catalog such as our heterocyclic compounds category before commissioning synthesis at all. A staple like 3-Iodopyridine (CAS 1120-90-7) delivered from stock in three days beats a custom quote every time.

Typical timeline: two to six weeks for a straightforward two- to four-step sequence. Anything quoted at under two weeks is either stock, a very short sequence, or optimistic.

Situation 2: Route Development for a Lead

You have a molecule that works and a route that does not scale: too many chromatographies, a cryogenic step at an awkward point, a 12 percent overall yield, a reagent that costs more than the product. You need a chemist to design a better path.

What to look for: process chemistry depth rather than catalog breadth. Ask who would own the route design, what their background is, and to see a redacted route-scouting report. This is the situation where contract R&D capability matters most and where the FTE commercial model usually fits. The build-versus-buy arithmetic for this stage is in contract R&D vs in-house labs: a cost analysis.

Typical timeline: eight to twenty weeks for a first-generation improved route, longer if the chemistry is genuinely novel.

Situation 3: Scale-Up for Tox or IND

You have a route and now need hundreds of grams to several kilograms with real documentation behind it. This is the situation where the CRO-to-CDMO boundary bites hardest, and where under-scoping is most expensive.

What to look for: demonstrated execution at your target scale within the last twelve months, a real safety assessment process for the steps that need one, and analytical capability that can generate a full impurity profile rather than an assay. Ask what happened the last time a scale-up failed at this scale and what they changed as a result. The answer is more informative than any capability list.

Typical timeline: twelve to thirty weeks depending on how much route work remains and whether the material supports a regulatory filing.

Situation 4: Ongoing Intermediate Supply

You need the same intermediate, repeatedly, at a predictable specification. The chemistry is settled. What you are now buying is reliability.

What to look for: at this point you should be screening on supply-chain criteria rather than chemistry criteria, meaning capacity, change control, second-source strategy, and documentation consistency lot to lot. Our chemical supplier qualification checklist covers the scoring detail, and pharmaceutical supply chain de-risking covers what happens when a single source fails.

Typical timeline: qualification of a second source for a non-GMP intermediate with an established method usually runs 60 to 90 days.

Situation 5: Analytical or Characterization Work Only

You have material and need it characterized, a method developed or transferred, or an impurity identified. No synthesis at all.

What to look for: instrument coverage and, more importantly, method development experience on your compound class. Anyone can run a sample. Developing a stability-indicating method that resolves a late-eluting degradant is a different skill. Ask for a redacted method development report. Our analytical services page lists the instrument set, and HPLC method development for APIs covers what that work actually involves.

Situation	Provider type	Commercial model	Screen hardest on
One-off screening compound	CRO or catalog supplier	Fee-for-service	Speed, stock position
Route development	CRO	FTE	Process chemistry depth
Tox / IND scale-up	CRO with scale, or CDMO	Milestone	Demonstrated scale, safety, analytics
Ongoing supply	CDMO or CMO	Per kilogram, contract	Capacity, change control, documentation
Analytical only	CRO analytical group	Fee-for-service or FTE	Method development experience

Most disappointments trace to a mismatch in this table rather than to poor chemistry. A CRO priced and staffed for Situation 2 will underperform on Situation 4, and it will underperform in ways that look like carelessness rather than like a structural mismatch.

Capability Screening: Telling a Capability From an Equipment List

An equipment list tells you what an organization bought. A capability tells you what it can execute reliably next month. The gap between the two is where most project failures originate.

Screen every claimed capability with the same three questions.

Question 1: What Comparable Chemistry Have You Run?

Ask for a description, in redacted form, of a project involving similar chemistry: similar bond formations, similar functional group tolerance, similar handling constraints. You are not asking for a customer name. You are asking whether the shape of the problem is familiar.

A good answer is specific and includes something that went wrong. “We ran a related Suzuki coupling on a chloropyridine at 400 grams; the first attempt stalled at 60 percent conversion and we solved it by switching the base and drying the solvent” is a real answer. “We have extensive cross-coupling experience” is not.

Question 2: Who Runs That Technique Today?

Equipment does not run itself, and specialist techniques concentrate in a small number of people. Ask how many chemists currently on staff routinely run the technique your route depends on. If the answer is one, ask what happens when that person is on leave. If the answer is evasive, the capability is on the website and not in the building.

Question 3: What Have You Run at My Scale in the Last Twelve Months?

The twelve-month qualifier matters. Capability decays. A hydrogenation vessel that has not been pressure-tested and run in two years is a project risk, not a capability. Ask for the largest scale executed on that specific technique in the last year, and ask what the scale ceiling is before the work would be subcontracted.

Subcontracting is not automatically disqualifying, and many good CROs subcontract a specialist step. Undisclosed subcontracting is disqualifying. Ask directly: which steps, if any, would you subcontract, and to whom, in what country. The domestic footprint picture is mapped in the US-based custom synthesis capability map.

The One-Page Capability Record

Keep the answers in a fixed format so that three providers can be compared side by side rather than remembered. For each critical technique, record: staffed operators, largest batch in the last twelve months, equipment specification, whether subcontracted, and one sentence on the comparable project cited. Five rows per provider is enough, and it converts a set of impressions into something you can defend to a program team.

The Specialist Chemistry Questions That Separate CROs

Most CROs can run a Suzuki coupling, an amide bond formation, and a Boc deprotection. The differentiation appears in the techniques that need dedicated equipment, dedicated training, or both. If your route touches any of these, the screening questions below matter more than anything else in this guide.

Cryogenic Chemistry

Work at −78 °C and below is common in organolithium and lithium-halogen exchange chemistry, and it is one of the clearest scale-up discontinuities. A dry-ice bath that works fine on 5 grams becomes a jacketed reactor with a chiller on 5 kilograms, and the cost profile changes completely. Ask what the largest cryogenic batch they have run is, what temperature they can hold and to what tolerance, and how they manage the exotherm on addition. The cost drivers are unpacked in cryogenic chemistry: what it costs and why.

High-Pressure Hydrogenation

Hydrogenation above a few bar requires rated vessels, a hydrogen supply with the right safety infrastructure, and operators trained on it. Ask for the pressure rating and working volume of the vessels, whether they are stirred autoclaves or something less capable, the catalyst handling process, and whether they can run a catalyst screen rather than only executing a known condition. The full question set is in high-pressure hydrogenation: the capability questions to ask.

Flow Chemistry

Continuous processing solves specific problems, including highly exothermic steps, unstable intermediates, and hazardous reagents generated and consumed in situ. It solves nothing at all for many others. The important screening question is not whether they have a flow rig but whether they can tell you when flow is the wrong answer. The comparison is in flow chemistry vs batch, and the regulatory dimension for anyone heading toward a filing is in ICH Q13 continuous manufacturing and CDMO readiness.

Fluorination

Fluorine chemistry spans a wide range of hazard and difficulty, from a benign nucleophilic displacement to reagents requiring specialist handling. Ask which fluorinating reagents they hold and handle routinely, and whether the answer covers your specific transformation. Many organizations that list fluorination handle building blocks only, buying in a fluorinated fragment from a catalog such as our fluorinated compounds category rather than installing fluorine themselves. That is often the right approach: a stocked intermediate like 4-Fluoroindole (CAS 387-43-9) is cheaper and faster than a custom fluorination. Just be clear about which one you are buying.

Chiral Synthesis and Resolution

Two questions separate real chiral capability from claimed chiral capability. First, can they develop a chiral analytical method rather than only run one you provide, because enantiomeric excess is meaningless without a validated separation. Second, what is their approach: asymmetric synthesis, classical resolution, chiral pool, or preparative chiral chromatography, and what determines the choice. A provider who only ever proposes one answer has one tool. Chiral building blocks such as (S)-2-Methylpyrrolidine (CAS 59335-84-1) are frequently a cheaper route to a single enantiomer than resolving late in a sequence.

Quality and Regulatory Posture

Match the quality level to the regulatory stage of the material, not to the most impressive certificate available. Paying for GMP you do not need is one of the most common and most avoidable cost overruns in chemistry outsourcing.

There are three practical levels, and the difference between them is documentation, not chemistry skill.

Level	What it means	Documentation produced	When you need it
Non-GMP	Research-grade, good scientific practice	Certificate of analysis, spectra, method summary	Discovery, SAR, in vitro, most preclinical
GMP-like	Non-GMP facility, enhanced documentation	Batch record, full analytical package, retained samples, traceable raw materials	Tox studies, IND-enabling work, some regulatory starting materials
Full GMP	Quality system, validated methods, audits	Master and executed batch records, validated methods, change control, deviation reports, formal QA release	Clinical and commercial API, material inside a filing

The Cost of GMP You Do Not Need

The chemistry in a GMP batch and a non-GMP batch is often identical. What differs is the quality system wrapped around it: pre-approved batch records, in-process controls with signed verification, validated rather than qualified analytical methods, full material traceability, deviation management, and a formal release step. That overhead is real work and it is priced accordingly. It also adds lead time, frequently several weeks, before any chemistry starts.

For a discovery compound going into an in vitro assay, none of it adds value. The distinction and its cost implications are covered in GMP vs non-GMP chemical manufacturing, and the buyer-side requirements when you genuinely do need the top level are in cGMP contract manufacturing requirements.

The Middle Level Is Usually the Right Answer for Preclinical

Most preclinical programs are best served by what the table calls GMP-like: a non-GMP facility producing a full documentation package. You get the batch record, the complete analytical characterization, the retained sample, and the raw material traceability that a tox package or a future regulatory starting material justification needs, without paying for a quality system your material does not yet require.

Ask explicitly for this level. Many CROs offer it and few advertise it, because it does not have a marketable acronym. The ICH quality guidelines define what the full level requires, which is a useful reference point for deciding how much of it you actually need at your stage.

How to Verify the Claim

A provider claiming GMP should be able to state which regulatory authority has inspected the facility and when, and to share the scope of that registration. A provider claiming GMP-like documentation should be able to show you a redacted batch record. Both are ordinary requests. Reluctance to answer either is a finding in itself. The FDA drug establishment registration database is the primary source for verifying a registration claim independently.

Commercial Structure: FTE, Fee-for-Service, or Milestone

Choose the commercial model based on how well-defined the deliverable is. A defined deliverable takes fee-for-service. An undefined one takes FTE. Everything in between takes milestones.

Fee-for-Service

You specify a compound, a quantity, and a specification; the provider quotes a fixed price. The risk of the chemistry not working sits with the provider, and they price that risk into the number. This is the right model when the target is known and the route is either known or straightforward.

Its weakness is scope change. If you alter the specification mid-project, you are renegotiating, and renegotiation mid-project rarely favors the buyer. Write the specification carefully up front. What to include in a custom synthesis RFQ lists the fields that prevent the most common disputes, and what drives your custom synthesis quote explains how the number is built, so you can tell an expensive quote from an overpriced one.

FTE

You buy a chemist, or a fraction of one, for a period. The provider carries no chemistry risk and you direct the work. This is the correct model for exploratory work: route scouting, medicinal chemistry campaigns, analog synthesis where the targets emerge from data as you go.

Its weakness is that it requires you to manage. An FTE arrangement with no technical oversight from the buyer produces activity rather than results. Budget your own time accordingly, typically two to four hours a week per FTE for review and direction.

Milestone

Payment tranches release on defined technical gates: route selected, gram-scale demonstrated, kilogram batch delivered to specification. This suits scale-up programs where each stage genuinely gates the next and where both parties want the risk shared rather than assigned to one side.

Its weakness is definitional. A milestone that is not objectively measurable becomes a dispute. Define each gate in terms a third party could adjudicate: a quantity, a purity by a named method, a documented yield over a defined sequence.

Rate Sanity Check

Rates vary widely by geography, technique, and regulatory level, and any single number quoted as a market rate should be treated skeptically. What is more useful is the shape of the comparison. A US-based FTE typically carries a multiple of an offshore FTE, and the real question is whether the difference is recovered in communication overhead, IP posture, time zone alignment, and rework avoided. For many discovery programs it is. For straightforward bulk supply it often is not, and pretending otherwise wastes budget that belongs elsewhere in the program.

IP and Confidentiality

Read the IP clauses before the pricing. A favorable price attached to an unfavorable IP position is not a good deal, and IP terms are far harder to renegotiate after work starts.

Background Versus Foreground IP

Background IP is what each party brings to the project: your compound, your target, your prior data; the CRO’s general synthetic know-how, its internal methods, its accumulated technique experience. Background stays with its owner. This is standard and should not be contentious.

Foreground IP is what arises during the project. The default position a buyer should seek is that foreground IP arising from work you commissioned, funded, and directed is assigned to you. Most reputable CROs accept this for standard custom synthesis, because their business is selling services rather than accumulating a patent estate on customer molecules.

The genuine grey zone is a general method improvement discovered while working on your molecule, such as a better catalyst system for a class of coupling, that has value well beyond your program. Some contracts assign this to the CRO with a license back to you. That can be acceptable. What is not acceptable is a clause that captures anything specific to your compound or your chemical series.

Clauses Worth Reading Twice

1. Scope of the CRO’s retained license. A license to use foreground IP for internal research is normal. A license to practice it commercially for other customers is not.
2. Publication rights. A CRO wanting to publish requires your prior written consent, without exception.
3. Residuals clauses. Language permitting the CRO to use information retained in the unaided memory of its personnel is broad and worth pushing back on for sensitive programs.
4. Definition of confidential information. It should cover the structure, the target, the data, and the existence of the project itself.
5. Survival period. Confidentiality obligations should survive termination by a defined number of years, and trade secret obligations should survive indefinitely.
6. Liability and reperformance. What happens if a batch fails specification. Reperformance at the provider’s cost is the normal remedy; a cap set at a fraction of the project fee is worth negotiating.

What a Reasonable NDA and MSA Look Like

A workable structure is a mutual NDA signed before any technical discussion, then a master services agreement carrying the IP, liability, and confidentiality terms, then individual statements of work per project referencing the MSA. This separates the terms you negotiate once from the scope you negotiate every time, and it is the reason experienced buyers can start a new project in days rather than weeks.

If a provider will not sign an NDA before a technical conversation, that is a process problem rather than a legal one, and it is usually fixable. If a provider’s MSA claims foreground IP by default and they will not amend it, that is a business model problem, and you should know it before you invest in the relationship.

Communication and Project Management

The strongest predictor of a good outcome is not the equipment, the certificate, or the price. It is the communication structure. Projects fail quietly far more often than they fail loudly, and the mechanism is almost always that bad news arrived too late to act on.

What a Good Weekly Update Contains

A useful weekly update is short and has five parts:

1. What was attempted this week, at the level of specific experiments rather than “work continued on step 3”
2. What the results were, with analytical data attached or referenced, not summarized in adjectives
3. What failed and the current hypothesis for why, which is the most important line and the one most often missing
4. What is planned for next week, specifically enough that you could object to it
5. Whether the timeline or budget has moved, stated explicitly even when the answer is no

An update that only ever contains parts 1 and 4 is a status report, not a technical update. Ask for the format above in writing at the start of the project. Providers who already work this way will agree immediately. Providers who do not will negotiate, which is itself informative.

The Named Chemist Test

Ask, before signing: who is the chemist who will run this, and can I speak to them for fifteen minutes before we start. This single question filters harder than any capability questionnaire. A provider who cannot name the person, or who insulates the technical staff behind account management entirely, is telling you how the project will be run.

You are not asking to bypass project management. You are asking whether technical judgment on your project is accessible when something goes wrong at 4 p.m. on a Thursday.

Escalation and Cadence

Agree three things at kickoff: the weekly update format and day, the maximum time a technical question can sit unanswered, and who is called when something goes badly wrong. Two business days is a reasonable response ceiling for a technical question. Anything longer and you will eventually find that a week of chemistry was run on an assumption you would have corrected in ten minutes.

Red Flags

Any single item below justifies another conversation. Two or more together justify removing the provider from your shortlist.

1. A quote returned with no clarifying questions. A non-trivial custom synthesis request almost always contains an ambiguity. A provider who quotes without asking either did not read it or is planning to resolve the ambiguity in their own favor later.
2. No named project chemist. Covered above. This is the single highest-value filter in the list.
3. Vague answers on analytical capability. “We have full analytical” is not an answer. The answer is a list of instruments, who operates them, whether they are in-house or subcontracted, and whether method development is included or charged separately.
4. Unwillingness to share a redacted example report. Every serious provider has example deliverables with customer information removed. Refusal usually means either no such reports exist or the quality would not survive review.
5. Pricing far below the market with no explanation. A quote at 40 percent of the field is a scope difference, not a discount. Find out which of purity, documentation, quantity, or analytical package has been quietly dropped. Sometimes the answer is legitimate, such as an existing intermediate on the shelf or a route they already know, and a good provider will tell you which.
6. No questions about your timeline or downstream use. A provider who does not ask what the material is for cannot tell you when the specification you requested is wrong.
7. Reluctance to discuss what could go wrong. Every route has a step most likely to fail. A chemist who cannot name it on your molecule has not yet thought about your molecule.

The Scorecard

Copy this into a spreadsheet, weight it for your situation, and score each shortlisted provider from 1 to 5. The weights below suit a preclinical scale-up. For a one-off screening compound, raise the weight on speed and lower the weight on quality system.

Criterion	Weight	What a 5 looks like
Demonstrated comparable chemistry	20	Describes a specific analogous project including what went wrong
Technique capability at your scale	15	Names a recent batch at or above your scale on the critical technique
Analytical capability and method development	15	In-house instruments, named operators, method development included
Quality posture matched to your stage	10	Offers the level you need and explains why more would be waste
Named technical contact and access	10	Chemist named and available for a pre-award call
Communication structure	10	Agrees to a defined weekly update format that includes failures
IP and contract terms	10	Foreground IP assigned, no residuals clause, mutual NDA before discussion
Commercial fit and transparency	5	Model matches the work, quote broken down by stage
Lead time and capacity confirmation	5	A specific start date rather than ‘we can start soon’

Score, multiply, sum. The output is rarely a surprise, but the exercise forces you to notice which criteria you were about to ignore. A provider that wins on price and loses on the first three rows is not the cheap option. It is the expensive option with a delayed invoice.

For a narrative version of the same decision, see how to choose a custom synthesis partner. External references worth reading alongside this guide: the FDA guidance documents for regulatory expectations at each stage, the ICH quality guidelines for documentation standards, and the ACS Green Chemistry Institute Pharmaceutical Roundtable for the solvent and reagent selection guides a good process chemist will already be using.

Before commissioning any synthesis, check whether the compound or a close precursor is already available. Our product catalog covers more than 7,000 compounds, and building blocks such as 4-Bromophenylboronic acid (CAS 5467-74-3) are frequently the shortest path to a target that looks like a custom project on first reading.

ChemContract Research has operated US-based chemistry from Huntington Beach, California since 2000. We provide custom synthesis from milligram to multi-ton, contract R&D for route development and optimization, and analytical services including HPLC, GC, NMR, LC-MS, ICP-OES, XRD, DSC/TGA, Karl Fischer, and chiral HPLC. Our synthesis capabilities include cryogenic chemistry to −78 °C, high-pressure hydrogenation, flow chemistry, fluorination, and chiral synthesis and resolution. Browse the full service list, or send us your specification and we will return a scoped quote, with a named project chemist, within 24 hours.

Frequently Asked Questions

What is the difference between a CRO, a CDMO, and a CMO?

A CRO provides research and development services: route scouting, method development, analysis, and small-quantity synthesis. A CDMO adds development plus manufacturing, typically under GMP, and carries your process into clinical or commercial supply. A CMO manufactures to a process you hand them and does not develop it. Mis-scoping between these three is the most common cause of an unusable quote.

How do I tell a real capability from an equipment list?

Ask three questions: what comparable chemistry have you run and can you describe it in redacted form, who on staff runs that technique today, and what is the largest scale you have executed on that technique in the last twelve months. An organization with the equipment but no current operator will quote it and then subcontract or fail. Demonstrated recent experience is the signal.

Do I need a GMP CRO for early discovery work?

Almost never. Discovery-stage compounds for screening, SAR, and in vitro work do not require GMP, and paying for GMP overhead at that stage adds cost and lead time without adding value. GMP becomes necessary when material enters a regulatory filing or is dosed in humans. Non-GMP with full analytical characterization is the correct posture for most preclinical work.

Which commercial model should I use, FTE or fee-for-service?

Use fee-for-service when the deliverable is defined: a compound, a quantity, a specification. Use FTE when the work is exploratory and the deliverable will change, such as route scouting or a multi-month medicinal chemistry campaign. Milestone structures fit scale-up programs where each stage gates the next. Mixing models across one project usually creates disputes about scope.

Who owns the route my CRO develops?

It depends entirely on the contract, so read it. A common and reasonable structure is that your background IP stays yours, foreground IP arising from work you paid for and directed is assigned to you, and the CRO retains its own pre-existing general know-how. Watch for clauses assigning improvements to the CRO or granting them a license to your compound class.

What are the clearest red flags in a CRO quote?

A quote returned with no clarifying questions, no named project chemist, vague answers on analytical capability, unwillingness to share a redacted example report, and a price far below the rest of the market with no explanation. Any one of these justifies a second call. Two or more together justify removing the vendor from the shortlist.