Hemolyzed specimens are rejected for coagulation testing based on concerns of artifactual interference. Prothrombin time (PT), activated partial thromboplastin time (aPTT), and selected factor assay test results for consecutive pairs of hemolyzed and subsequently recollected (mean, 72 minutes later) nonhemolyzed patient specimens were compared. Specimens from healthy human subjects were subjected to mechanically induced hemolysis, and PT and aPTT results compared with concurrently drawn nonhemolyzed control samples. In 50 paired patient specimens, there were statistically significant differences in PT (15.8 +/- 8.4 vs 16.3 +/- 8.7 seconds, P < .01) and aPTT (31.6 +/- 18 vs 32.5 +/- 19 seconds, P < .01) between hemolyzed and nonhemolyzed specimens, respectively. Specimens from healthy subjects showed no difference in PT and a minor difference in aPTT. A policy of rejecting hemolyzed specimens for coagulation tests should be revisited because the observed difference, when present, is unlikely to be considered clinically meaningful.