Participant-level randomized trial benchmark

RCT Bench

A curated public collection of randomized controlled trials for evaluating covariate-adjustment methods.

Explore metadata Download workbooks

125 Trials

-- Participant rows

-- Cleaned variables

-- Publication span

Benchmark contract

Cleaned trials, publication-aligned metadata, direct file access.

Each trial is stored as a participant-level CSV and RDS table with a shared naming contract for treatment, primary outcomes, secondary outcomes, and baseline covariates.

The explorer below is generated from the public metadata and data dictionary workbooks, keeping the web view synchronized with the benchmark files.

Metadata lookup

Find a trial

Loading trials...

Search Research area Primary outcome type Randomization Sort

Trial	Paper	Area	Sample	Outcome	Files

Downloads

Workbooks and trial files

Metadata workbook Trial-level publication metadata and provenance sheets. Data dictionary One row per cleaned public variable. All data and dictionaries Download every cleaned CSV and RDS trial file with all per-trial data dictionaries.

Using RCT Bench

Please acknowledge the work behind the data.

If you use this resource in your research, please cite the primary publication corresponding to each trial you use, as well as the RCT Bench methods paper:

Shao, Y., Lyu, L., Yu, M., & Wang, B. (2026). How should covariates be handled in randomized trials? Empirical evidence from 50 trials and recommendations for practice. Journal of Clinical Epidemiology, 112374.

Curation

Built for covariate-adjustment research

Eligibility

Individually randomized trials with participant-level treatment assignment and at least one outcome.

Variable contract

Treatment, YP_*, YS_*, and X_* columns keep analysis inputs consistent.

Publication alignment

Primary outcomes, registry identifiers, paper links, and citation counts are tied to the primary publication where available.