BACKGROUND: Selecting elderly people with colorectal cancer (CRC) for adjuvant chemotherapy is challenging. Comprehensive geriatric assessment (CGA) can help by classifying them according to their frailty profile. The supposed benefit of chemotherapy is on the basis of the rate of treatment adherence. In this study we evaluated tolerance and adherence to tailored-dose adjuvant therapy on the basis of CGA in a cohort of older patients with high-risk stage II and stage III CRC. PATIENTS AND METHODS: This was a prospective study in 193 consecutive patients aged 75 years or older. On the basis of CGA results, we classified patients as fit, medium fit, or unfit, administering standard therapy, adjusted treatment, and best supportive care, respectively. We recorded planned chemotherapy, toxicity, and completion of the treatment. A logistic multivariate analysis was carried out. RESULTS: Seventeen (15%) of the 141 candidates for chemotherapy (n = 86 fit and n = 55 medium fit) refused treatment; associated factors included polypharmacy (odds ratio [OR], 5.34; 95% confidence interval [CI], 1.55-18.40) and rectal location (OR, 5.61; 94% CI, 1.45-21.49). Of the 105 (74%) patients receiving chemotherapy, 20 (27%) fit and 4 (13%) medium fit patients experienced Grade 3 to 4 toxicity (P = .11) without association to explanatory variables. Approximately 55% of patients treated with chemotherapy received at least 80% of the planned dose (55% fit and 58% medium fit patients; P = .7). Factors associated with completion of chemotherapy were the absence of toxicity (OR, 7.67; 95% CI, 2.41-24.43) and social support (OR, 2.29; 95% CI, 0.08-1.04). CONCLUSION: CGA is useful for selecting elderly patients for adjuvant chemotherapy, adapting the dose to their frailty profile, and identifying adherence-related factors amenable to modification through CGA-based interventions.