The study involved the random assignment of eligible offenders between ISP and routine probation. Of 14 sites participating in the ISP demonstration, three programs in Los Angeles, Ventura, and Contra Costa counties were selected for separate study. Primary data sources were official record data collected for individual probationers, contextual information regarding program implementation, and criminal justice cost data for each site. The study looked at the degree to which ISP components were actually delivered and the extent to which delivered ISP components differed from those actually provided in routine probation programs. All three ISP programs were more intensive than routine probation. At each site, ISP participants had significantly more face-to-face and telephone contacts, as well as drug tests, than their routine counterparts. In terms of recidivism, between 41 and 73 percent of all studied probationers had new technical violations. For ISP offenders, about 33 percent had a new arrest. No statistically significant differences were observed between ISP and routine probationers in the severity of arrest offenses at any of the three sites. The analysis did not find ISP to be more effective than routine probation in reducing recidivism overall nor to be differentially effective for offenders with varying background characteristics. In addition, ISP was more costly than routine probation. 63 footnotes, 11 tables, and 6 figures