The application of quantitative pathology to the study of malignant tumors is discussed, with emphasis on the correlation between nuclear roundness and prognosis in prostatic carcinoma. Stage A2 prostatic carcinoma patients with 4-year follow-ups and surgically treated stage B1 and B2 prostatic carcinoma patients with 15-year follow-ups were characterized by manual digitization of cancerous nuclei. A form factor that assessed nuclear shape produced an accurate separation of those patients who developed metastatic disease from those who did not. The digitized nuclear roundness correlated with the perpendicular light scatter measured by flow cytometry. When combined with forward light scatter, perpendicular light scatter separated four tumor cell lines of different metastatic potential in an animal model of prostatic cancer; a relative grading index based on these findings seemed capable of identifying human patients with more extensive disease in a preliminary study. The observation of freshly aspirated viable prostatic carcinoma cells led us to appreciate membrane ruffling, pseudopodal extension and translational movements that appear to distinguish these cells from their benign counterparts. Time-lapse cinematography, image digitization and Fourier analysis allow for the objective description of cell motion and should provide new and exciting tools for the pathologist.