Data from the Bureau of Justice Statistics project on the prosecution of felony arrests for 1980 are analyzed. These data cover 14 jurisdictions, all with populations of 200,000 or more. The median ratio of pleas to trials among these jurisdictions is 11 pleas for every trial; however, 3 jurisdictions have more than 20 pleas for every trial, and 3 others have only 4 or 5 pleas per trial. Data suggest that high crime rates and the pressure of large case loads may not sufficiently explain the variation. Jurisdictions with a high proportion of trials differ from those with a high proportion of pleas in several respects. They are more selective in screening arrests, rejecting cases at higher rates than the high-plea-rate jurisdictions. Additionally, high-trial-rate jurisdictions appear less likely to reduce the principal charge when accepting guilty pleas than are high-plea-rate jurisdictions. Although they may produce fewer incarcerations per arrest, the high-trial-rate jurisdictions tend to produce slightly more long-term imprisonments per arrest and more per conviction than the low-trial-rate jurisdictions. Several related research issues are identified. Six data tables are included.