To evaluate the effect of sampling method and cytometric method on DNA ploidy results, a comparison study was performed on 20 whole prostate glands removed at prostatectomy. Fresh sampling was by sampling fine-needle aspiration (FNA). Paraffin sampling was by microdissection and re-embedding of 3 to 13 (average 6) 5-mm foci of microscopically proven tumor. Analysis was by flow cytometry and by image cytometry (microscopically guided). In tumor negative cases, flow and image cytometry of the FNA was diploid in each case, and flow cytometry of paraffin-embedded tissue was diploid in 5/6 cases. In tumor positive cases, non-diploid tumor was detected by image cytometry of the FNA in 70%, by flow cytometry of the FNA in 29%, and by flow cytometry of paraffin-extracted nuclei in 21%. The most effective combination was sampling fine-needle aspiration and image analysis.