We read with interest the commentary by Eidelman et al1 in regard to aspirin, postmenopausal hormones, and C-reactive protein (CRP). The authors discuss the potential value of CRP, a sensitive marker of inflammation, in predicting cardiovascular events and the effects of aspirin in reducing CRP. There are 2 additional points worth considering: first, how aspirin produces the anti-inflammatory effect and, second, that "endogenous" salicylic acid may influence the background inflammatory state and the effects of aspirin. Aspirin has a short half-life and is rapidly metabolized and hydrolyzed to salicylic acid, its principal metabolite.2 The anti-inflammatory action of aspirin is probably due to salicylic acid,3 although how salicylic acid exerts its anti-inflammatory action in vivo is the subject of much research; salicylic acid has little effect on cyclooxygenase-2 activity in vitro. There is substantial interindividual variation in serum salicylic acid concentrations after the administration of aspirin, 75 mg/d.4 This is related, perhaps, to differences in salicylic acid metabolism and renal excretion.2 Such variation in serum salicylic acid concentrations may cause significant differences in the anti-inflammatory effects for a given aspirin dosage. Salicylic acid is present also in the serum of individuals not consuming salicylate drugs,4- 5 probably arising from fruits and vegetables in our diet, and there is overlap in the serum concentrations of salicylic acid found in vegetarians and patients taking 75 mg of aspirin daily.4 The serum salicylic acid concentrations present in some individuals not taking aspirin may therefore be sufficient to produce an equivalent anti-inflammatory effect to low-dosage aspirin and affect the background inflammatory state. If such individuals take aspirin as well, then the presumed anti-inflammatory effects of the aspirin may be overstated.