Existing research suggests that incentives to plead guilty may influence guilty and innocent defendants differently. This study examined that possibility through testing theoretically informed predictions relating to the interaction between different types of plea discount (sentence length and sentence type), guilt, and probability of conviction in predicting plea decisions, with a focus on discounts in England and Wales. Participants (N = 3,375) made plea decisions in vignettes that varied discount type offered, probability of conviction, and guilt between subjects. Participants also answered questions about considerations that were important to them when making plea decisions. Results provide support for predictions, specifically at higher levels of probability of conviction, by showing that a discount resulting in a categorically different sentence type (probation rather than custody) encouraged both “guilty” and “innocent” participants to plead guilty but that a discount resulting only in a shorter sentence of the same type (a one-third reduction in sentence length) only encouraged “guilty” participants to plead guilty. Participant reports of the considerations important to them when pleading suggest that the categorical discount reduced the importance of factual guilt or innocence in the decision-making of innocent defendants. Findings suggest that utilizing plea discounts that vary sentences quantitatively but not categorically is important in maximizing the extent to which plea discounts appeal to guilty but not innocent defendants. (PsycInfo Database Record (c) 2022 APA, all rights reserved)