Cost of Running a Clinical Trial: Budgeting

Search interest in “cost of randomized controlled trials” is usually driven by one of two very different questions: a policy audience asking what drug development costs society, or a research administrator who actually has to build, negotiate, and defend a trial budget line by line. This guide answers the second question. It is written for the person assembling or reviewing a clinical trial budget — a sponsored-programs officer, clinical trials office administrator, or coordinating-center manager — not a general “how expensive is a clinical trial” explainer.

Two things make trial budgeting hard to generalize: the cost categories themselves, and the fact that per-patient and per-study costs vary by an order of magnitude depending on trial phase and therapeutic area. Below, both are addressed directly — first the line items a budget actually needs to cover, then what the published cost-estimation literature says about the range and its drivers, with real studies cited rather than a single average figure presented as fact.

Why clinical trial costs vary so much

There is no single “cost of a clinical trial” figure that means anything without qualification, for three structural reasons a budget builder needs to account for from the outset.

Phase

Cost scales with the size, duration, and monitoring intensity of the trial, all of which change by FDA phase. Early-phase trials (Phase 0/1) enroll small numbers of participants over short periods, often in a single unit, with lighter monitoring burden. Phase 3 trials — the pivotal, typically randomized, controlled trials that support a marketing application — are larger, longer, multi-site, and carry the heaviest data-quality and safety-reporting requirements, which is reflected directly in cost. Published per-day and per-study cost estimates (cited below) consistently show Phase 3 as the most expensive phase per trial, even though Phase 2 sometimes exceeds it on a per-patient basis due to smaller, harder-to-recruit populations.

Therapeutic area

Recruitment difficulty, protocol complexity, visit frequency, and the specialized procedures a protocol requires all differ by disease area, and this shows up as a large, consistent spread across every published cost study. Oncology and rare-disease trials tend to run expensive because eligible populations are small and hard to find, follow-up is long, and correlative/biomarker testing adds procedure cost. Cardiovascular and other trials that require large sample sizes to detect a clinically meaningful effect on a hard outcome (e.g., mortality, major adverse cardiac events) also run expensive, but for a different reason — sheer patient volume and multi-year follow-up rather than recruitment difficulty per patient.

Trial design

Design choices that don’t show up in a topline “average cost” figure move the actual budget substantially: whether the trial is placebo-controlled or active-comparator-controlled (an active comparator typically costs more — procurement, matching, and often a higher required sample size), whether the primary endpoint is a hard clinical outcome versus a surrogate (clinical-outcome trials require longer follow-up and larger samples), and total enrollment target. These design variables are set during protocol development, which is why a research administrator should be involved in budget modeling before the design is finalized, not after.

What a research administrator needs to budget for

A complete clinical trial budget is built from several distinct cost categories, each negotiated or estimated differently. Missing one of these during initial budget development is a common source of mid-trial budget shortfalls.

Per-patient and site costs

The core of most industry-sponsored trial budgets is a per-patient (or “per-subject”) rate negotiated with each site, covering the protocol-required procedures, visits, and staff time for one enrolled participant who completes the study per protocol. Sites typically negotiate this rate against a standard fee schedule, sometimes with separate line items for screen failures (a lower rate, since fewer procedures are completed) and early withdrawals. Site startup costs — IRB submission preparation, contract negotiation, staff training, pharmacy/investigational-product setup — are usually a separate fixed fee, not amortized into the per-patient rate, because they’re incurred regardless of how many patients a site ultimately enrolls.

IRB and ethics review fees

Institutional Review Board or Research Ethics Committee review carries its own fee, which varies depending on whether the trial uses a local institutional IRB or a central/commercial IRB (commonly used for multi-site industry trials to avoid duplicative site-by-site review). Continuing review, amendment review, and unanticipated-problem/deviation review each typically carry incremental fees on top of the initial review fee, and a realistic budget should plan for at least one or two amendment cycles over a multi-year trial rather than assuming the protocol will never change after initial approval.

Data management and CTMS

Electronic data capture (EDC) build and hosting, database design and validation, query management, and the ongoing licensing or per-record cost of a Clinical Trial Management System (CTMS) used to track enrollment, visit compliance, and site performance are a distinct, often underestimated, line item. See CASRAI’s clinical data management guide for the standards (CDISC data standards, 21 CFR Part 11 electronic-records requirements) that shape what a compliant data management build actually costs.

Monitoring

Monitoring visit costs — travel, Clinical Research Associate (CRA) time, and the source-data-verification effort required at each visit — scale with monitoring frequency and model. Traditional 100% source-data-verification monitoring costs more per site-visit than a risk-based monitoring (RBM) approach, which concentrates in-person verification on higher-risk sites/data points and relies more heavily on centralized statistical monitoring between visits. ICH E6(R2)’s explicit endorsement of risk-based, proportionate monitoring is one of the reasons RBM adoption has grown — it is a genuine cost lever, not just a compliance nicety, and a budget built on a fixed monitoring-visit cadence without considering a risk-based alternative may be overbuilt for a low-risk protocol.

Regulatory and CRO fees

If the trial is managed in whole or in part by a Contract Research Organization (CRO), the CRO’s fee — covering project management, regulatory affairs support, medical monitoring, and often the monitoring function itself — is typically the largest single non-site line item in a sponsor’s budget. CRO fee structures range from fixed-bid to time-and-materials to hybrid models, each shifting financial risk differently between sponsor and CRO; a research administrator reviewing a proposed CRO budget should confirm which model is being used and what happens to the fee if enrollment runs faster or slower than planned.

Participant payments

Compensation and reimbursement paid directly to enrolled participants — for time, travel, and burden — is a distinct budget line from site fees and CRO fees, governed by separate ethical (IRB undue-influence review) and tax-reporting rules. See CASRAI’s dedicated guide, Clinical Trial Participant Payments: Structure, Ethics, and IRB Review, for payment-structure options (per-visit, prorated, lump-sum) and the current IRS 1099 reporting threshold. Don’t fold this line item into the per-patient site rate in a budget model — sites usually pass it through to participants directly and track it separately for tax purposes.

Indirect costs (F&A)

For federally-funded and many institutionally-hosted trials, the site’s negotiated Facilities & Administrative (F&A) rate applies on top of direct costs, typically calculated against a Modified Total Direct Cost (MTDC) base rather than the full budget — see CASRAI’s guide to Uniform Guidance (2 CFR 200), the governing framework for how indirect costs are negotiated and applied on federal awards. Industry-sponsored trials handle this differently: some institutions apply a separate, often higher, negotiated clinical-trial overhead rate rather than the federally negotiated F&A rate, since industry funding isn’t subject to 2 CFR 200. Confirm which rate applies before finalizing a budget — using the wrong one is a common and consequential error.

What the published cost-estimation research actually shows

Rather than quote a single headline number, it’s more useful for budget planning to know what the peer-reviewed and federally-commissioned cost literature has actually found, and how widely the estimates vary depending on scope and methodology. Three studies are the most frequently cited primary sources in this space:

Sertkaya et al. (2014), for HHS/ASPE

“Examination of Clinical Trial Costs and Barriers for Drug Development,” prepared by Eastern Research Group for the U.S. Department of Health and Human Services’ Office of the Assistant Secretary for Planning and Evaluation, analyzed negotiated clinical trial contracts and protocols (Medidata Solutions data, compiled from 2004 onward) to estimate average per-study costs by phase and therapeutic area. It reported average per-study costs of roughly $3.8 million (Phase 1), $13.35 million (Phase 2), $19.89 million (Phase 3), and $19.95 million (Phase 4) across therapeutic areas — with substantial variation around those averages by therapeutic area (for example, Phase 3 pain and anesthesia trials averaged roughly $52.9 million, well above the cross-area Phase 3 average). Read the full report at aspe.hhs.gov.

Moore, Zhang, Anderson & Alexander (2018), JAMA Internal Medicine

“Estimated Costs of Pivotal Trials for Novel Therapeutic Agents Approved by the US Food and Drug Administration, 2015-2016” examined 138 pivotal trials supporting 59 new FDA drug approvals and found a median trial cost of $19.0 million (interquartile range $12.2 million–$33.1 million) — but with an overall range spanning less than $5 million to $346.8 million, illustrating just how little a single average communicates. Cost per patient had a median of roughly $41,117. Trial characteristics moved cost substantially: uncontrolled trials averaged $13.5 million versus $48.9 million for active-comparator-controlled trials, and trials measuring a hard clinical outcome averaged $64.7 million versus $24.0 million for those using a surrogate endpoint. By therapeutic area, cardiovascular trials in this sample averaged $157.2 million, well above oncology ($45.4 million) and infectious disease ($22.1 million) trials in the same dataset. Published in JAMA Internal Medicine, 178(11), 1451–1457.

DiMasi, Grabowski & Hansen (2016), Journal of Health Economics

The Tufts Center for the Study of Drug Development’s widely-cited “Innovation in the Pharmaceutical Industry: New Estimates of R&D Costs” is a full-development-lifecycle estimate, not a per-trial figure, and it’s important not to conflate the two. Based on data from 106 randomly selected investigational compounds across 10 pharmaceutical firms, the study estimated average out-of-pocket cost per approved new compound at $1.395 billion (2013 dollars), rising to a capitalized total (including the cost of capital tied up during the 10+ years of development) of roughly $2.6 billion. This figure covers preclinical research and all clinical phases combined, plus the cost of the many candidate compounds that fail during development and never reach approval — it is not comparable to the single-trial figures above, and citing it as “the cost of a clinical trial” is a common but incorrect simplification worth correcting when it comes up in budget conversations with non-specialist stakeholders.

Taken together, these three sources make the same structural point from different angles: there is no single defensible “average cost of a clinical trial” figure, and any budget built around one industry-wide number instead of a bottom-up build from the categories above is likely to be wrong in a specific, correctable way.

Routine patient-care costs and Medicare coverage

A budgeting question that comes up specifically for trials enrolling Medicare beneficiaries: Medicare covers the routine (non-investigational) patient-care costs associated with a qualifying clinical trial under National Coverage Determination (NCD) 310.1 — a separate, unrelated definition from NIH’s or FDA’s definition of a clinical trial. Qualifying criteria include that the trial’s subject or purpose falls within a Medicare benefit category, has therapeutic intent, and meets one of several deeming criteria (including certain federal funding sources or conduct under an FDA IND). This coverage determination affects how routine care costs are allocated between the trial budget and standard clinical billing — a coverage analysis, distinguishing routine costs from research-only costs, is a standard step for any trial billing Medicare, and getting it wrong creates real compliance exposure. See CMS’s own NCD 310.1 page at cms.gov for the full qualifying criteria.

Building the budget: a practical framework

For a research administrator assembling or reviewing a trial budget, the categories above translate into a practical build order:

Start from the protocol, not a template. Visit schedule, procedures per visit, and total enrollment target drive per-patient cost directly — a budget built from a generic template before the protocol is finalized will need substantial revision.
Separate fixed costs from variable costs. Site startup fees, IRB initial review, and CTMS/EDC build are largely fixed regardless of enrollment; per-patient fees, participant payments, and monitoring visits scale with actual enrollment and retention.
Confirm the applicable indirect cost rate before finalizing. Federally-funded trials use the institution’s negotiated F&A rate against an MTDC base under 2 CFR 200; industry-sponsored trials may use a separate negotiated clinical-trial overhead rate instead.
Budget for amendments. At least one or two protocol amendments — each triggering IRB re-review fees and potential CRO change-order costs — over a multi-year trial is realistic, not a worst-case assumption.
Build in contingency for enrollment risk. Slower-than-planned enrollment extends the timeline over which fixed costs (site retention, CTMS licensing, staff salaries) are incurred, even if the per-patient budget itself doesn’t change.
Confirm Medicare/payer coverage analysis is complete for any trial enrolling Medicare beneficiaries or billing routine care to a third-party payer, before finalizing the participant/site cost model.

Frequently asked questions

How much does it cost to run a clinical trial?

There is no single reliable figure — published per-study cost estimates range from roughly $3.8 million (average Phase 1) to well over $150 million for large, complex Phase 3 trials in expensive therapeutic areas, depending on phase, therapeutic area, trial design, and geography. See “What the published cost-estimation research actually shows” above for the specific studies and ranges behind that variation.

What is the average cost per patient in a clinical trial?

One widely-cited estimate, from a JAMA Internal Medicine analysis of 138 FDA pivotal trials (2015–2016 approvals), found a median cost per patient of roughly $41,117, with substantial variation by therapeutic area and trial design (active-comparator and hard-outcome trials cost more per patient than placebo-controlled or surrogate-endpoint trials).

Why do clinical trial costs vary so much by therapeutic area?

Recruitment difficulty, protocol complexity, procedure intensity, and required follow-up duration differ substantially by disease area. Oncology and rare-disease trials tend to be expensive because eligible patients are hard to find and follow-up is long; cardiovascular trials tend to be expensive because they require very large sample sizes to detect an effect on a hard outcome like mortality.

What’s the difference between the Tufts $2.6 billion figure and a single trial’s cost?

The Tufts Center for the Study of Drug Development’s widely-cited figure is a capitalized, full-development-lifecycle estimate — covering preclinical research, all clinical phases, and the cost of failed candidate compounds, averaged across an entire drug development program. It is not the cost of any single clinical trial and should not be quoted as one.

Does Medicare cover the cost of a clinical trial?

Medicare covers the routine (non-investigational) patient-care costs of a qualifying clinical trial as defined under CMS National Coverage Determination 310.1 — a separate definition from NIH’s or FDA’s definition of a clinical trial. It does not cover the investigational item or service itself, or costs incurred solely for research purposes. A coverage analysis is needed to separate routine costs from research-only costs in the trial budget.

What clinical trial budget line items are most often underestimated?

Protocol amendment costs (IRB re-review fees plus CRO change orders), the ongoing licensing/data-management cost of a CTMS across the full trial duration, and the extension of fixed costs caused by slower-than-planned enrollment are common sources of mid-trial budget shortfalls.

The Cost of Running a Clinical Trial: A Budgeting Guide for Research Administrators