We report a study in which our objective was to analyze the clinical response during IAD in patients with biochemical failure after RRP for clinically localized prostate cancer. Between February 1994 and May 1996, 34 patients who exhibited a primary postoperative decrease in PSA to below the detection limit after RRP and then showed PSA progression during follow-up were included as group 1 and 17 patients in whom PSA did not decrease after RRP were included as group 2. Patients were offered IAD when PSA progressed over 0.4 ng/ml in group 1 and over 4.0 ng/ml in group 2. Median follow-up is 184 weeks in group 1 and 206 weeks in group 2. The median time "off" therapy increased from 25% (1st cycle) to 68.7% (5th cycle) of the entire cycle in group 1 and from 33.3% to 58.3% in group 2. Nine out of 12 cases with Gleason score > or =8 failed to respond to IAD and all developed metastatic and/or local failure. No case with Gleason score <7 failed to respond to IAD. Our conclusions suggest that IAD may be effective in patients with biochemical progression after RRP. In our experience, Gleason score seems to be an important variable.