Objective To demonstrate regression to the mean bias introduced by matching on preperiod variables in difference‐in‐differences studies. Data Sources Simulated data. Study Design We performed a Monte Carlo simulation to estimate the effect of a placebo intervention on simulated longitudinal data …