Nuclear shape analysis performed upon prostatectomy specimens of prostatic carcinoma distinguished individual patients with good and poor prognoses. In order to be useful for prognosis assessment preoperatively, nuclear morphometry must be measured on needle biopsy specimens. We compared nuclear morphometry on automatic biopsy and radical prostatectomy specimens in 20 patients with prostatic carcinoma. Nuclear size was smaller (paired Student t test, P < .0001) in biopsy specimens (perimeter 17.0 +/- SD 4.9 microns, area 29.9 +/- 6.6 microns2) than in prostatectomy specimens (perimeter 24.6 +/- 4.4 microns, area 48.2 +/- 8.7 microns2). Nuclear shape was more abnormal in automatic biopsy specimens (nuclear roundness factor 82.0 +/- 18.8, ellipticity 90.5 +/- 27.7) than in surgical specimens (nuclear roundness factor 43.5 +/- 8.8, ellipticity 54.0 +/- 14.7) (P < .0001). Study of specimens obtained by automatic biopsy preoperatively and automatic biopsy of the prostatectomy specimens at various steps of processing revealed that nuclear swelling and rounding occurred after 2-24 hours of formalin fixation. Automatic prostate biopsies may more accurately reflect true nuclear morphometry and should be studied for preoperative prognosis prediction in patients with clinically localized prostatic carcinoma.