It is widely assumed that criminal defendants who face multiple charges in a single trial have a harder time prevailing than those who face several trials of one count each. Conventional wisdom also has it that a defendant who is joined for trial with other suspects is worse off than one who stands trial alone. Until now, these assumptions have never been tested empirically. Looking at nearly 20,000 federal criminal trials over a five-year period, this Article asks if the traditional beliefs are true, and if so, tries to measure the impact on trial outcomes of joining counts and defendants. It turns out that joinder has a significant prejudicial effect on trials, although not quite the same effect that is usually assumed. Using statistical models that control for a range. of variables, the authors discovered that trial defendants who face multiple counts are roughly 10% more likely to be convicted of the most serious charge than a defendant who stands trial on a single count. Surprisingly, however, joining co-defendants in a single trial had virtually no impact on the likelihood of conviction, at least in the aggregate. Using the results of the empirical study, the Article then reconsiders the competing policy issues that underlie the joinder and severance doctrine, and explores the implications of the findings.