The long-term effectiveness of the Parent-Child Development Centers (PCDCs) as programs to prevent behavior problems in children was examined with follow-up data collected 6-13 years after program completion. Data were collected for 581 children who had been in the programs with their mothers (Ns: Birmingham, 151; New Orleans, 186; Houston, 244). Mothers and teachers were interviewed. There were few significant differences between program and control groups. Only the early cohorts of the Houston program showed significant differences between groups on the Child Behavior Checklist (CBCL). EDITORS' STRATEGIC IMPLICATIONS: This is a rare example of long-term longitudinal evaluation of a cross-site prevention program with a large sample size. Practitioners and program designers will be interested in the author's descriptions of cohort and site implementation differences. The absence of major effects at follow-up (despite significant short-term effects) in this well-designed study must caution us against thinking of early prevention programs as inoculations.