The objective of this study was to evaluate needle biopsy of recurrent prostate cancer after radical prostatectomy. We evaluated 37 cases of recurrent prostate cancer after radical prostatectomy that were diagnosed by needle biopsy between March 1984 and July 1998. Fifteen were from consultations in which contributors were uncertain of the diagnosis, and 22 were from men who had come to The Johns Hopkins Hospital for treatment. The median interval from radical prostatectomy to biopsy showing recurrent tumor was 40 months. There was no correlation between the interval to recurrence and either pathologic features of the biopsy and radical prostatectomy or various clinical features. The mean extent of adenocarcinoma in the biopsies was 3.2 mm (range, 0.1 to 18 mm; median, 2 mm). The length of recurrent cancer on biopsy correlated with an abnormal rectal examination (P = .001). The mean Gleason score for the recurrent tumors was 6.5, which correlated with the grade of the radical prostatectomy cancer (P = .005). The cancers often lacked overt histologic features of malignancy. Benign prostatic acini were seen in five cases (14%), usually separate from the cancer. In 5 (33%) of the consultation cases, we would not have been able to diagnose cancer if not for the fact that atypical prostate glands should not be present after radical prostatectomy. In well-sampled radical prostatectomies, margins were almost always positive, as was extraprostatic extension. In cases with less sampling, there was a higher incidence of organ-confined disease and margin-negative disease implying suboptimal processing of the radical prostatectomy. After radical prostatectomy, recurrent cancer on needle biopsies may be focal and difficult to diagnose and must be assessed differently than in patients who have not had surgery.