There is limited data in the literature that suggests that transition zone (TZ) biopsy might be useful for the prediction of extraprostatic extension (EPE) in clinically localized prostate cancer. We studied the role of TZ biopsy in the prediction of EPE. Transition zone biopsies were performed in addition to systematic peripheral zone (PZ) biopsies between November 1995 and December 1999. During this period, 59 patients underwent radical prostatectomy for clinically localized disease. Final pathological results were compared with preoperative clinical and biopsy findings. Of the 59 patients who underwent radical prostatectomy, 46 had cancer only in the PZ cores and 13 had cancer both in the PZ and the TZ cores at the biopsy. Final histopathological results revealed EPE in 19 (32%) patients and positive surgical margins in 22 (37%). In univariate analysis of age, prostate-specific antigen (PSA), mean percentage of positive PZ cores, mean biopsy Gleason score and positive TZ biopsy, there was a significant difference for serum PSA levels (P = 0.021), presence of positive TZ cores (P = 0.018) and percentage of positive PZ cores in patients with and without EPE (P < 0.001). In multivariate analysis, the single independent predictor of EPE was the percentage of positive PZ biopsy cores (P = 0.0227). There was agreement between the side of positive TZ biopsy and EPE in seven of eight patients. Taking two TZ cores in addition to peripheral sextant biopsy did not result in better prediction of EPE. The relationship between TZ involvement and the presence of EPE can be investigated further in radical prostatectomy specimens.