Biotech due diligence should verify whether the company’s central scientific claim survives independent review and whether the requested capital can reach a defined value-changing milestone. Before funding, you should connect the science, clinical plan, regulatory record, intellectual property, manufacturing, market, team, and cash needs into one evidence chain.
Pitch decks often separate those topics into clean slides, but investment risk moves across them. A promising mechanism can lose value when the assay can’t be repeated, the trial endpoint doesn’t support the proposed claim, patent ownership is unclear, or the production process can’t supply comparable clinical material. This checklist helps you test what’s true, what remains assumed, and what must happen before the company needs more capital.
Check 1: Reduce the Company Story to a Testable Scientific Claim
Start by rewriting the investment thesis as one testable sentence. It should name the biological target or pathway, the therapeutic modality, the intended patient group, the proposed mechanism, and the measurable result expected from treatment. If the sentence depends on broad phrases like “platform potential” or “disease modification,” ask management to replace them with defined biological events and decision-ready measurements. You can’t evaluate a claim until each part can be supported or disproved.
Separate evidence that a target is associated with disease from evidence that changing the target can improve a meaningful outcome. Review human genetics, patient samples, natural-history data, target-expression patterns, pharmacology, and independent replication where they apply. Search PubMed for the original papers, later studies, corrections, competing findings, and work from groups with no financial relationship to the company. NIH RePORTER can also help you trace publicly funded projects, investigators, and linked research outputs rather than relying only on the company’s selected bibliography.
Ask for the full evidence package behind the key figure in the pitch deck: protocol, raw data, inclusion and exclusion rules, controls, replicate counts, analysis code, image files, and explanations for removed observations. Check whether experiments were randomized or blinded when those methods fit the study, and whether the reported result survives reasonable alternative analyses. A reproducibility guide published through the National Library of Medicine stresses transparent methods, sound design, quality control, and appropriate statistical analysis as recurring requirements in preclinical research. Your goal is to identify which result has been independently reproduced and which result still rests on one team, one assay, or one batch.
Check 2: Audit Whether Preclinical Evidence Can Translate to People
Preclinical diligence should test whether the program has earned the proposed human dose, patient population, and clinical hypothesis. Review the relevance of each model, the relationship between dose and exposure, pharmacokinetics, pharmacodynamic markers, durability, safety findings, immunogenicity, and the link between the measured signal and the intended clinical benefit. FDA describes preclinical research as the stage where laboratory and animal testing addresses basic safety questions before clinical research begins. Treat a positive efficacy chart as one part of that package, not as the package itself.
Request final study reports and raw files from internal laboratories and contract research organizations. Reconcile the company’s summary with every completed study, including inconclusive experiments, failed replications, unexpected toxicities, assay changes, and studies stopped early. Verify that the test article used in key studies matches the intended clinical candidate in sequence, formulation, delivery system, impurity profile, and manufacturing history. A result generated with research-grade material can’t automatically support assumptions about later clinical-grade material.
Build a translation table with four columns: preclinical finding, human prediction, clinical measurement, and failure signal. A tumor-growth result may predict a defined response measure, yet the translation also depends on exposure, target engagement, disease biology, prior treatment, and patient selection. Set a funding condition when a central translation gap can be answered before closing or through a staged investment. Money should follow evidence gates that management can measure, not calendar dates chosen for fundraising.
Check 3: Read the Clinical Protocol Instead of the Trial Headline
Clinical diligence begins with the protocol, statistical analysis plan, registry record, and patient-level data available for review. Verify the population, control group, randomization, masking, dose selection, primary endpoint, follow-up period, missing-data rules, multiplicity plan, safety monitoring, and prespecified subgroup analyses. Then compare those documents with the slide deck and public statements. A “positive study” can still carry major risk when the claimed benefit came from an exploratory endpoint, a post hoc subgroup, or a protocol change that wasn’t explained.
ClinicalTrials.gov study records identify the sponsor, eligibility criteria, study design, outcome measures, locations, and record dates. Posted results can include participant flow, baseline characteristics, outcome measures, and adverse events, including serious adverse events. Compare the current record with earlier versions to see when endpoints, enrollment targets, eligibility rules, status, or completion dates changed. Registry data won’t replace source-data review, but it can expose questions that a polished presentation leaves unanswered.
Examine whether the endpoint measures direct clinical benefit or acts as a biomarker or surrogate. FDA defines a surrogate endpoint as a substitute for a direct measure of how a patient feels, functions, or survives, and the agency maintains a table of surrogate endpoints used or considered for drug and biologic approval. Check the exact disease, treatment setting, endpoint definition, measurement method, and proposed regulatory use rather than assuming acceptance transfers across indications. When the program depends on a novel biomarker, ask what evidence links movement in that marker to the claimed patient benefit.
Check 4: Verify the Regulatory Record From Original Documents
Ask for original agency correspondence rather than a management summary saying the meeting was “positive.” Review meeting requests, briefing packages, written responses, official minutes, clinical-hold communications, designation letters, protocol feedback, and the company’s follow-up commitments. FDA’s formal-meetings guidance covers sponsor interactions related to drug and biological-product development and review, so the exact questions asked and the agency’s exact responses matter. A favorable comment on one issue does not settle an issue that the company never presented.
Separate an expedited designation from marketing authorization. FDA states that Breakthrough Therapy designation is designed to speed development and review, and agency policy says the designation is not drug approval and does not change the approval standards for safety and effectiveness. Verify Fast Track, Breakthrough Therapy, regenerative-medicine, orphan, and other claimed statuses from the actual designation letter or an official public database. Do not add approval probability merely because a program offers earlier or more frequent agency interaction.
Check public claims against FDA’s approval and designation resources. Drugs@FDA covers most approved prescription, generic, and over-the-counter drugs; the Purple Book lists licensed biological products; the orphan-drug database identifies public designation and approval records. FDA also explains that orphan designation is a separate process from approval or licensure. Record each unresolved agency concern, the evidence needed to answer it, the planned interaction, and the cash and time required if the answer is unfavorable.
Check 5: Test Patent Ownership, Claim Scope, and Freedom to Operate
Confirm who owns each patent and application, what the issued claims cover, which jurisdictions matter, when protection may end, and what obligations sit inside every license. Review inventor assignments, employee agreements, university licenses, sponsored-research agreements, collaboration terms, prosecution files, maintenance status, and recorded ownership changes. The USPTO Patent Public Search tool provides access to published patents and applications, and its Assignment Search database contains recorded patent-assignment information available to the public from August 1980 onward. Use those records to verify the data room, not as a substitute for a legal opinion.
Owning a patent does not by itself establish freedom to operate. Patent counsel should compare the planned product, manufacturing method, delivery system, patient-selection method, and commercial activities against live third-party claims in each target market. Ask for claim charts, prior-art analysis, prosecution history, written-description and enablement review, opposition or litigation history, and a clear statement of known blocking rights. Pending applications deserve separate treatment because their claims can change, narrow, or fail to issue.
Read license economics line by line. Verify upfront and milestone obligations, royalties, sublicensing shares, diligence duties, field and territory limits, patent-control rights, termination rights, change-of-control clauses, and rights retained by universities or research partners. Check whether background inventions and later improvements belong to the company, a founder, an academic institution, or a collaborator. Align the realistic patent and exclusivity period with the remaining development schedule; protection that fades soon after launch supports a different valuation than a slide showing the earliest filing date.
Check 6: Inspect Manufacturing Readiness Before It Becomes the Bottleneck
Manufacturing diligence should determine whether the company can repeatedly produce material with the intended identity, strength, quality, purity, potency, and safety profile. Review the process description, control strategy, release specifications, analytical methods, stability program, batch history, deviations, out-of-specification investigations, corrective actions, and comparability plans. FDA’s current good manufacturing practice rules set minimum requirements for the methods, facilities, and controls used to manufacture, process, and pack drug products. A clinical program can advance only as fast as its supply and quality systems allow.
Trace the production chain from starting material to released dose. Identify single-source reagents, cell banks, plasmids, viral vectors, resins, specialized assays, fill-finish slots, cold-chain needs, import dependencies, and long-lead equipment. Review quality agreements and master service agreements with contract manufacturers, then confirm reserved capacity, technology-transfer status, batch timing, yields, and responsibility for deviations. A signed manufacturing contract means little when the main slot, method transfer, or raw material remains unsecured.
Compare the material used across discovery, toxicology, early clinical work, and the planned registration program. Process changes can alter impurities, potency, stability, delivery, and observed clinical performance, so the company needs a documented plan to establish comparability. Ask which assays are qualified or validated at the current stage, which specifications remain provisional, and which process parameters have limited operating data. Put cost and schedule ranges around failed batches, repeat testing, delayed release, additional stability work, and a change of manufacturer.
Check 7: Rebuild the Market From the Treatable Patient Up
A market model should begin with patients who can be identified, reached, treated, and paid for under the proposed label. Start with disease prevalence or incidence, then narrow for diagnosis rate, disease stage, biomarker status, prior therapy, contraindications, treatment setting, physician capacity, site readiness, and expected eligibility criteria. Compare each assumption with disease registries, peer-reviewed epidemiology, clinical-practice guidance, claims data, and interviews with treating specialists. A large disease population does not automatically create a large launch population.
Map approved and investigational competition by mechanism, treatment line, endpoint, safety, dosing, administration burden, manufacturing demands, and likely launch timing. ClinicalTrials.gov can identify registered studies and their status; Drugs@FDA and the Purple Book can confirm approved drug and biologic products. Read approved labels and study designs rather than grouping every product aimed at the same disease into one competitor list. A therapy can show biological differentiation and still struggle commercially when another product is easier to prescribe, administer, monitor, manufacture, or reimburse.
Test the revenue model against the full treatment pathway. Verify who orders the test, who confirms eligibility, where treatment occurs, how long onboarding takes, what monitoring is required, who carries inventory, and which party absorbs treatment failures or delays. Review pricing research, payer feedback, coding and payment plans, site economics, patient out-of-pocket exposure, and expected discounts. The company’s addressable market should shrink when real access barriers are added; a model that never narrows is a warning sign.
Check 8: Verify That the Team Can Execute the Claimed Plan
Evaluate the team against the work required to reach the upcoming financing milestone. A discovery-stage program needs scientific leadership, yet a clinical-stage company also needs trial operations, statistics, regulatory, safety, manufacturing, quality, vendor management, and financial control. Verify employment history, publication authorship, patent inventorship, prior program responsibilities, agency interactions, and the outcomes of earlier programs. PubMed, patent databases, trial registries, former colleagues, board members, and service providers can help confirm what each executive actually led.
Look for hidden dependence on one founder, academic laboratory, consultant, contract manufacturer, or clinical investigator. Review employment and consulting agreements, invention-assignment terms, retention plans, succession coverage, scientific-advisory commitments, conflicts of interest, and related-party transactions. Ask who owns each major decision and whether that person has the authority, time, and operating support to deliver it. A respected name on a slide adds limited value when the person attends one meeting a quarter and holds no execution responsibility.
Run reference calls around decisions, not personality. Ask how management handled negative data, protocol problems, manufacturing deviations, missed enrollment, budget pressure, and disagreement with outside experts. Compare those accounts with board materials, internal forecasts, staff turnover, vendor history, and changes in development strategy. You’re testing whether leaders surface bad news early and revise decisions from evidence before cash is lost.
Check 9: Rebuild the Financing Need From the Work Plan
Recalculate the funding request from the underlying development plan rather than accepting the company’s monthly burn summary. Verify cash, debt, payables, committed purchase orders, lease obligations, license milestones, trial contracts, manufacturing deposits, headcount plans, insurance, quality-system costs, and legal spending. Tie each major cost to a contract, quote, enrollment assumption, batch plan, or hiring schedule. Separate committed spending from management estimates and separate milestone-related work from optional expansion.
The financing plan should reach a defined decision point: a reproducible experiment, regulatory interaction, clinical readout, manufacturing release, or other event that can change program value. Build base, delay, enrollment-shortfall, manufacturing-failure, and negative-data cases, then identify when cash falls below the level needed to continue safely. Include time for data cleaning, database lock, analysis, agency preparation, and fundraising rather than ending the runway on the last patient visit. A program that reaches an event but lacks cash to interpret or defend it has not truly reached the milestone.
Review the capitalization table and all rights that can alter ownership or future financing. Confirm issued shares, options, warrants, convertible instruments, founder vesting, liquidation preferences, participation rights, pro rata rights, anti-dilution terms, information rights, board rights, and contractual limits on a new round. Model dilution across the current financing and at least one later round under several valuations. Your investment memo should state which risks are accepted, which require a price adjustment, and which must be resolved before funds are released.
What Should Investors Check Before Funding a Biotech Company?
- Scientific reproducibility
- Preclinical and clinical data
- FDA path and correspondence
- Patent ownership and freedom to operate
- Manufacturing readiness
- Market, team, milestones, and cash runway
Fund the Evidence Chain, Not the Story
A biotech investment becomes easier to judge when every claim leads to original evidence, every development step leads to a measurable gate, and every dollar leads to a defined milestone. Verify the science first, then test whether the clinical design, regulatory record, patents, production plan, market, and team support the same thesis. Record uncertainties rather than smoothing them into one probability, since different failures demand different remedies, timelines, and financing terms. Use staged funding, closing conditions, or price changes when a material question can be answered before you assume the full risk. The best diligence does not eliminate uncertainty; it shows you exactly which uncertainty you are funding and what evidence will change your decision.
References
- U.S. National Library of Medicine. About PubMed
- National Institutes of Health. NIH RePORTER
- National Library of Medicine. Improving the Quality and Reproducibility of Preclinical Research
- U.S. Food and Drug Administration. The Drug Development Process
- ClinicalTrials.gov. How to Read a Study Record
- U.S. Food and Drug Administration. Surrogate Endpoint Resources for Drug and Biologic Development
- U.S. Food and Drug Administration. Formal Meetings Between the FDA and Sponsors or Applicants of PDUFA Products
- U.S. Food and Drug Administration. Breakthrough Therapy
- U.S. Food and Drug Administration. Drug Approvals and Databases
- U.S. Food and Drug Administration. Drugs@FDA
- U.S. Food and Drug Administration. Purple Book Database of Licensed Biological Products
- U.S. Food and Drug Administration. Orphan Drug Designations and Approvals
- United States Patent and Trademark Office. Patent Public Search
- United States Patent and Trademark Office. Patent Assignment Search
- U.S. Food and Drug Administration. Current Good Manufacturing Practice Regulations