FDA's own review clock is flat-to-slightly-faster over 20 years, but overall drug development time has not shortened

Summary

Over the last twenty years, the FDA's own review clock has NOT gotten dramatically faster or slower — it plateaued after a sharp drop that occurred in the 1990s/early-2000s following the Prescription Drug User Fee Act (PDUFA, 1992). Comparing 2005 to 2022 application cohorts: standard-review approval time fell modestly, from a median of about 13.1 months to 10.0 months (about 24% faster). Priority-review approval time actually got a bit slower, from about 6.0 months to 8.0 months (about 33% longer), though it remains far faster than standard review. Meanwhile the FDA has approved roughly twice as many novel drugs per year (about 25/year in the 2000s vs. about 50/year in 2018-2024), driven by heavier use of expedited pathways. But the part of the process BEFORE FDA review — clinical development — has gotten somewhat LONGER, not shorter (Tufts CSDD: 83.1 months average for 2008-2013 approvals vs. 89.8 months for 2014-2018 approvals). So the honest answer depends on which clock you mean: FDA's own decision time is flat-to-modestly-improved; the full bench-to-approval pipeline that a drug actually experiences is roughly unchanged at 10-15 years, if anything slightly longer on the clinical-trial side.

What the AskAmerica corpus could and couldn't do

health.fda_drug_approvals (openFDA-sourced) carries only the final approval status and date — per its own documented column description, openFDA's drugsfda feed carries no separate application-submission/filing date anywhere in the record. That means review time (submission-to-approval duration) is not computable from this connector's own tables. This is a genuine data gap, confirmed via describe_table and search_catalog, not a query mistake. Accordingly, every number in this report comes from FDA's own primary PDUFA Performance Reports to Congress, fetched and read directly, plus one industry-standard clinical-development benchmark (Tufts Center for the Study of Drug Development) for the pre-FDA-review portion of the timeline.

FDA's own review/approval time, 2005-2022

FDA publishes, cohort by year of application receipt, the median time from receipt to approval for two tracks: priority review (serious conditions, target ~6 months) and standard review (target ~10 months). Pulling the values directly from FDA's own tables:

Standard review shows a modest net improvement (13.1 → 10.0 months, -24%) but with no steady trend in between — it actually rose back to ~12 months for several years (2013-2020) before falling again in 2021-2022. Priority review shows a modest net lengthening (6.0 → 8.0 months, +33%), which FDA itself partly attributes to procedural changes such as the Program for Enhanced Review Transparency and Communication (which added review-clock time starting FY2013) and to more first-in-class/complex applications using the priority track. Neither track shows the kind of dramatic multi-decade acceleration that occurred between 1993 (26.6-month median pre-PDUFA) and the mid-1990s/2000s; that big drop happened BEFORE the twenty-year window this question asks about, and the last twenty years have been a plateau by comparison.

More drugs are being approved, largely via expedited pathways

Even as the per-application review clock has been roughly flat, the number of novel drugs approved per year has grown substantially: about 25 per year in 2000-2010, about 34 per year in 2011-2017, and about 50 per year in 2018-2024 (CDER novel drug approval counts). In 2025, 74% of the year's 50 novel approvals used at least one expedited regulatory pathway (fast track, breakthrough therapy, priority review, and/or accelerated approval). A separate peer-reviewed analysis of all 1980-2022 FDA approvals (Seoane-Vazquez et al., Scientific Reports 2024) corroborates this: the median FDA review time fell from 26.6 months before PDUFA (1992) to 9.9 months after the 2012 FDASIA reforms, and more than half of new-drug approvals since then have used a pathway that did not require demonstrating added therapeutic benefit over existing treatments — i.e., speed has been bought partly by lowering the evidentiary bar for a given application, not solely by FDA working faster on a fixed bar.

The part before FDA review has gotten longer, not shorter

"Time to get a drug approved" is ambiguous between FDA's own review clock and the full pipeline a drug goes through, most of which happens before FDA ever sees the application. Per Tufts Center for the Study of Drug Development, drugs approved in 2014-2018 spent an average of 89.8 months in clinical trials, up from 83.1 months for drugs approved in 2008-2013 — a roughly 8% increase over that decade, moving in the opposite direction from FDA's own (flat-to-improving) review clock. Combined with preclinical research, total bench-to-approval time for a drug has remained in the broad 10-15 year range across the last twenty years, essentially unchanged in the big picture despite the faster FDA decision component, because slower/larger clinical trials have offset the FDA-side gains.