The purpose of this study was to assess whether regimen comprehension deteriorates in the elderly without obvious mental disability. Eligible patients were elderly who could visit hospitals by themselves. We recruited 138 patients (age: 43-89, 75 males and 63 females, underlying diseases: hypertension, hyperlipidemia, arrhythmia etc.) from our outpatient clinic. The participants were tested with a regimen comprehension scale (RCS: Jpn J Geriat 1997; 34:209-214). The differences in scores among individuals increased with age. Scores of 5 or less were recorded in 10 of 69 patients aged 65 or more, but in non of 69 patients aged less than 65 (p < 0.01). The 60 patients who could not get full marks were classified into 2 groups: the tutored group (T) who were tutored by pharmacists about taking drugs, and the non-tutored group (N). Both group were tested again with RCS to evaluate the effect of tutorial. In T-group (n = 28), the second scores increased significantly (from 7.2 +/- 0.9 to 8.6 +/- 2.0 (m +/- SD); p < 0.01). Although the second scores showed a tendency to increase in N-group (n = 29), there was no statistical significance. In 7 patients who obtained less than a score of 5 on the RCS and age- and gender-matched controls who received full marks on the RCS, the HDS-R test failed to show any differences between the two groups. Thus, we concluded that even in the self-attending elderly patients, the regimen comprehension deteriorated with age, and tutorials were considered to be effective.