Abstract Background: The prevalence and course of pseudothrombocytopenia in outpatients is uncertain. Methods: In a cohort study of 687,955 members of a health maintenance organization, we extracted 36,780 consecutive automated complete blood count test results and determined the point prevalence of pseudothrombocytopenia during a one-month period. We also calculated a retrospective cumulative prevalence over the past 5 years. Results: There were 1105 (2.7%) patients with platelet counts of 100–149×109/L and 304 (0.8%) with counts <100×109/L, of whom 12.8% (n=39) had pseudothrombocytopenia newly discovered or in the past. There were 40 additional patients with a history of pseudothrombocytopenia, but now showing platelet counts of 150×109/L or more, and another 21 with platelet counts of 100–149×109/L. The total cumulative prevalence of pseudothrombocytopenia was 0.27% (100/36,780), 2.5-fold higher than the point prevalence of 0.11% (39/36,780). Platelet counts were extremely variable in these patients during the 5-year follow-up period. Conclusions: Physicians should be aware of the variability of platelet counts in patients with pseudothrombocytopenia in order to reduce unnecessary retesting. Since pseudothrombocytopenia can result in platelet counts between 100 and 149×109/L, peripheral smears should be done in all patients with such values, or the reference limit value should be lowered to 100×109/L.