Cumulative Illness Rating Scale (CIRS) remains difficult to use in older patients, especially because of a possible memory bias while declaring a chronic disease among patients with cognitive disorders. Because acute and chronic diseases are usually treated with drugs, we hypothesized that the number of drug classes taken per day could be a surrogate measure of comorbidity burden and, thus, could be positively associated with the CIRS-G score. The aim of this study was to determine whether the CIRS-G score was associated with the number of drug classes taken per day by older inpatients in a geriatric acute care unit. Based on cross-sectional design, 324 older inpatients (85.3 ± 6.4 years, 63.3% female) were prospectively included in this study. Number of drug classes daily taken was recorded using the Anatomical Therapeutic Chemical Classification (ATCC) and the CIRS-G score was also calculated. Among studied older inpatients, the mean CIRS-G score was 8.6 ± 3.6 and the mean number of drug classes daily taken was 7.0 ± 3.7. The linear regressions showed that only the number of drug classes daily taken was significantly and positively associated with the CIRS-G score (coefficient of regression β = 0.317 for unadjusted model, β = 0.304 for fully adjusted model and β = 0.317 for backward model with all P-values < 0.001). Our findings show that there is a direct association between the CIRS-G score and the number of drugs classes daily taken among the studied sample of older inpatients.