Purpose: To identify a cut-off score for the COmprehensive Score for financial Toxicity (COST) to predict a clinical implication of a high level of financial toxicity (FT). Methods: A total of 640 cancer patients were recruited from three regional hospitals in Hong Kong. They completed a questionnaire comprising the COST measure and the Functional Assessment of Cancer Therapy - General (FACT-G) instrument. The cut-off score for the COST that predicts the lowest quartile of the FACT-G total score was identified by receiver operating characteristic (ROC) analysis. The sample was then stratified by this cut-off score, and characteristics were compared using Fisher’s exact, chi-squared or independent sample t-test. Results: The mean scores were 20.1 ± 8.8 for the COST and 71.6 ± 15.5 for the FACT-G. The ROC analysis suggested that the cut-off of 17.5 yielded an acceptable sensitivity and specificity. Characteristics of patients with a higher level of FT included being younger, having a monthly household income of < 10,000 HKD (approximately 1290 USD), being more likely not employed, having stage IV cancer and receiving targeted and/or immunotherapy. In terms of financial support, a higher proportion of these patients had discussed financial issues with health care professionals and had received financial assistance. In addition, fewer of them were covered by private health insurance. Conclusion: Our findings suggest a cut-off for the COST that can be used to screen for FT in clinical settings. In addition, while a considerable proportion of high-FT patients received targeted therapy, they often received financial assistance. There is a gap between financial hardship and assistance that warrants attention.