BACKGROUND: This study aimed to investigate the effects of different frailty dimensions on frailty prevalence in older Taiwanese cancer patients receiving chemotherapy, and to analyze the dimensions that should be included in frailty assessment for effectively predicting serious adverse events, unexpected hospitalizations, and emergency department visits. MATERIALS AND METHODS: This study prospectively enrolled 234 cancer patients with solid cancer or lymphoma and aged 65 years or older who later received chemotherapy at a medical center in Taiwan from September 2016 to November 2018. First, all patients were subjected to a frailty assessment on eight frailty dimensions within 1 week before their first chemotherapy treatment. The effects of different dimensions on frailty were analyzed using a Poisson regression model. Second, after sequentially excluding one, two, and three dimensions with the lowest effects, frailty was sequentially assessed in the remaining seven, six, and five dimensions for comparison of chemotherapy-related adverse events. RESULTS: Nutritional status, comorbidity, history of falls, cognitive status, and polypharmacy were the top five important dimensions of frailty in older Taiwanese cancer patients. Regardless of the number (five to eight) of dimensions used for frailty assessment, frail patients had higher rates of serious adverse events, unexpected hospitalizations, and emergency room visits than non-frail patients during chemotherapy. CONCLUSIONS: Frailty assessment in older Taiwanese cancer patients should be based on at least five dimensions to accurately identify those at high risk of serious adverse events during chemotherapy. It is expected that the present findings may be used to design a frailty scale for older Taiwanese in the future.