The study finds that few felons in the sample were good candidates for probation, and calls for a greater array of sentencing options that should be restrictive enough to ensure public safety. During the 40-month followup period, 65 percent of the probationers were rearrested, 51 percent were reconvicted, 18 percent were reconvicted of serious violent crimes, and 34 percent were given a jail or a prison sentence. A high correlation was found between being sentenced to prison and factors such as drug addiction. For all offenses except assault, offenders with three or more of the relevant characteristics had an 80-percent probability of going to prison. A statistical analysis of offenders' sentences revealed that many of those granted felony probation were indistinguishable -- in terms of their crimes and criminal records -- from those imprisoned. A statistical model of good prospects for probation was applied to a larger sample of prisoners sentenced in 1980 to determine which offenders would have a good chance of succeeding on probation; only about 3 percent of this population was identified as having at least a 75-percent chance of success. The report recommends alternative, intermediate forms of punishment (such as intensive surveillance), which the researchers suggest can restore probation's credibility and reduce imprisonment rates without increasing crime. Nine figures, 3 appendixes, and approximately 100 references are supplied.