The Rey-Osterreith Complex Figure (ROCF) is commonly used in evaluations of patients undergoing epilepsy surgery. We assessed test-retest performance on ROCF in 30 partial epilepsy patients (mean interval = 33.7 months) to derive reliable change indices (RCIs) and regression-based measures for change. ROCF reproductions were rescored by three raters (IRR Copy: 0.963; Delayed Recall: 0.986). The derived adjusted RC (90% CI) cutoff values for the ROCF Copy were (or=8.4) and were (or=10.0) for the Delayed Recall. Results from regression-based analyses were negative, using age, education, seizure duration, and age of onset, whereas a baseline score was a significant predictor of a follow-up score. The results provide a means to evaluate long-term outcome in epilepsy patients using the ROCF.