Purpose We sought to investigate key factors that legal scholars believe are contributing to false guilty pleas. Methods Undergraduates playing the role of an innocent defendant in a plea bargaining scenario (N = 382) were randomly assigned to condition in a 2 (trial penalty: mandatory minimum of 10 years, judicial discretion of 4–6 years) × 3 (likelihood of conviction: 10%, 50%, 90%), × 2 (attorney recommendation: take the plea deal, use best judgement) between‐subjects design – false guilty pleas (2 years in prison) were the primary dependent variable. Results Participants faced with a mandatory minimum were more likely to falsely plead guilty, and this effect was mediated by perceptions that life would be more difficult serving the sentence associated with losing at trial compared to that of the plea. Participants given a 90% likelihood of conviction at trial were more likely to falsely plead guilty than those in the 50% and 10% conditions, and this effect was mediated by perceptions of case strength. Finally, female participants (via increased perceptions of case strength) were more likely to falsely plead guilty than male participants. Contrary to our predictions, attorney recommendation did not impact false guilty pleas, perhaps due to the absence of a proper social influence component to the manipulation. Conclusions Overall, the findings provide evidence supporting the apparent causal mechanisms underlying the ‘innocence problem’ in plea bargaining, and we discuss future directions for plea bargaining research pertaining to both research questions and methodology.