Research rarely has shown that in-program change in correctional rehabilitation is related to long-term outcome (i.e., recidivism), and surprisingly little is known about what happens to progress after treatment, especially for "lifers" whose release may not be imminent. This study investigated patterns of treatment response for 35 life-sentenced treatment completers of an intensive cognitive-behavioral program for high-risk male violent prisoners. Using Violence Risk Scale (VRS) ratings at pre-treatment, post-treatment, and 6 to 12 months following the program, we found that prisoners’ mean treatment response was positive both at program end and follow-up. However, a fine-grained analysis identified five distinct change patterns within the sample. Importantly, the direction and volume of in-program change did not necessarily predict post-program change, and the highest risk prisoners did not benefit as much as those at medium-high risk. The findings suggest that (a) a better understanding of the effects of treatment may be gained by examining change beyond the end of interventions, including a focus on the individual and contextual factors that promote and inhibit generalization and (b) more therapeutic attention may be warranted for monitoring treatment change to maximize conditions for continued gain beyond the end of the program.