BACKGROUND. The value of rectal examination as initial screening test for prostate cancer at low PSA values (0.0-3.9 ng/ml) was determined by evaluating the number and tumor characteristics of the cancers detected. METHODS. Two study populations were subjected to screening with (n=10,226) and without (n=10,753) rectal examination as initial screening test. The number of cancers detected at low PSA values for both screening regimens, the corresponding biopsy and radical prostatectomy tumor characteristics were assessed. Possibly harmless cancers were defined as small (<0.5 ml) organ-confined tumors without Gleason growth-patterns 4/5. RESULTS. At low PSA, 26.6% (117/440) of screen-detected cancers were detected after the evaluation of a suspicious rectal examination. The number of cancers and tumor aggressiveness features were highly associated with serum-PSA level. The proportion of possibly harmless disease steadily declined from 100% (PSA 0.0-0.9 ng/ml) to 15.4% (PSA 3.0-3.9 ng/ml). Rectal examinations were performed unnecessarily in 94.7-100% of cases, when detection of clinically significant disease was aimed at. Using PSA (and a cut-off of 3.0 ng/ml) as the only screening tool, 24.3% (121/498) of screen-detected cancers were in the PSA range 3.0-3.9 ng/ml, and 60.0% were assessed as clinically significant. CONCLUSIONS. Rectal examination as initial screening test for prostate cancer at low PSA values may be replaced by screening using serum-PSA only. At PSA levels below 3.0 ng/ml, 289 rectal examinations are required to find one case of clinically significant disease, and 96 rectal examinations are needed to diagnose prostate cancer of any size, grade, or stage.