Data were drawn from court and police records of 1,378 defendants charged with felonies in 12 North Carolina counties during 3 months in 1979. Defendants in the sample were mostly male, with a median age of 23. Almost half were black. The likelihood that all the defendants' felony charges would be dismissed and sentence severity if charges were not dismissed were correlated with the cases' basis characteristics, such as availability of an eyewitness, an incriminating statement, and seriousness of the charge. Likelihood of dismissal and sentence severity were also associated with administrative variables independent of the basic case factors. Defendants charged with a Class 2 felony (larceny or breaking and entering) were less likely to have charges dismissed if they had assigned counsel than privately paid counsel. If charges were not dismissed, the defendants with assigned counsel were more likely to receive severe sentences than if represented by privately paid counsel. The more time a defendant spent in detention, the lower the odds of dismissal and the more severe the sentence if convicted. One result of these associations was that black defendants were placed at a disadvantage because they were more likely than white defendants to have assigned counsel and spend more time in pretrial detention. Defendants' prospects for dispositions varied substantially among the 12 counties, reflecting differences in prosecutorial practices. Plea bargaining was extensive, and sentences received by those convicted at trial were much more severe than for those who pleaded not guilty. The defendant's decision to go to trial was not associated significantly with intrinsic characteristics of the case, suggesting that administrative variation in plea bargaining was responsible for many plea/trial decisions that defendants made. Ways to reduce variation and channel discretion in interim decisions are suggested. Tables and 68 footnotes are supplied.