Variations in the absorption pattern of aspirin in enteric-coated tablets and in solution were compared between subjects and between doses in any one subject. Absorption of enteric-coated tablets was complete but the rate was extremely variable. On the other hand, administration of aspirin in solution was not only complete, but also yielded well-reproducible salicylate plasma levels. The variable absorption of drugs in enteric-coated tablets appears to be due to wide differences in gastric retention time of solid objects. Since this aspect of gastrointestinal physiology is not affected by the type and quality of the tablets or the medicinal agent, it constitutes an inherent limitation in the absorption efficacy of all enteric-coated tablets. There was no significant difference in the metabolic fate of aspirin administered in solution or in the enteric-coated tablets used in this study.