The influence on confession evidence in trials is quite strong; triers of fact who hear confession evidence find these self-incriminating statements hard to ignore and in turn, vote to convict more often. However, most cases do not see the inside of a courtroom, but rather are resolved via plea bargains. In the present study, we examined how confessions, whether partial or full, influence guilty plea rates and plea discounts (the difference between sentence received at trial if convicted and sentence received as part of the plea). We coded more than 500 district attorney case files for defendant statement type (i.e., not questioned by police, questioned but denied guilt, questioned and partially confessed, questioned and fully confessed), case disposition (guilty plea, trial, dismissal), and other pertinent information (e.g., initial charges, perceived strength of evidence). We found that whereas those who denied guilt were the least likely to plead guilty, when they did plead, they enjoyed the largest plea discounts. In addition, partial and full confessors were found to be equally likely to plead guilty (both at near-ceiling levels), but partial confessors received the smallest plea discounts by far. Our findings have implications for theories of remorse and punishment, and plea decision-making.