The DNA content of 369 rectal cancers was measured by flow cytometry. One hundred and four (28%) were diploid, 252 (68%) were aneuploid, and 13 (3.5%) were tetraploid. Diploid cancers were associated with an improved 5 year survival (p less than 0.001) and were more likely to present at an early stage. DNA content, however, did not confer independent prognostic information in a Cox model based on four discrete pathological variables. Patients were classified by a new system of prognostic grouping and those with a very good or a very poor outlook were removed leaving 137 prognostic group III patients. No further substratification of this group by DNA content or by four additional pathological variables could be achieved. As the new prognostic system is not improved by the addition of ploidy, routine adoption of flow cytometry in the assessment of rectal cancer cannot be recommended.