Marshall (1997) assessed the effectiveness of the Grendon regime by examining 4 years of reconviction data on all inmates who were selected for Grendon in the years 1984 to 1989. The Marshall study also compared the reconviction rates of those inmates selected to Grendon against a small group of offenders who had been selected to Grendon but did not actually go there. The current study used the same sample and method as that used by Marshall. The sample of just over 700 inmates who had been admitted to Grendon in the period 1984 to 1989 is called the "admitted group." A separate analysis of "lifers" is included in the current study. The study found that prisoners selected for Grendon tended to be high-risk offenders when compared with similar inmates from the general prison population. This could be due to the fact that Grendon selects inmates with personality disorders. Lower rates of reconviction were found for inmates who went to Grendon than for inmates selected for Grendon but who did not go there. Time spent at Grendon was strongly related to reconviction; reconviction rates were lower for inmates who stayed for at least 18 months. There was evidence of the effect of treatment on reconviction rates for Grendon life sentence inmates, probably due to a longer time in treatment. There apparently is some reduction in the reconviction rate for violent offenses among the treatment group and for sexual and violent offenses among repeat sexual offenders. 3 figures, 2 tables, and 3 references