A major concern of physicians caring for patients is the development of hyperkalemia, a potentially life-threatening event requiring accurate determination of its etiology. After metabolic and iatrogenic causes have been excluded, factitious hyperkalemia must be considered, one cause of which may be the method of laboratory determination. Hyperkalemia associated with thrombocytosis has been previously described but is a fact commonly overlooked in the evaluation of a patient with hyperkalemia. We compared the potassium levels in simultaneous serum and plasma samples from patients with normal and elevated platelet counts, since platelet activation during clot formation is associated with release of potassium. We found consistently higher potassium levels in serum (containing products of activated platelets) than in plasma (containing nonactivated platelets). The greatest discrepancy in potassium levels was in patients with thrombocytosis. This preliminary study suggests that if there is no obvious explanation for an elevated serum potassium the plasma potassium level should be measured, particularly in patients with elevated platelet counts.