Introduction: Early detection of malnourishment in hospitalized patients is essential. Poor eating is a risk factor for malnourishment and worsening the prognosis. Therefore, a simple and accurate method to assess food intake should be considered. This study aimed to validate a tool evaluating patients’ food acceptance from a major hospital in São Paulo, Brazil. Methods: This cross-sectional study included 60 patients of both sexes (73.2±1.8 years old, 56.7% male) at risk for malnourishment. Patients/caregivers filled out the food acceptance evaluation tool. An evaluator verified the food left on trays after meals and compared their responses with those of the patients’. The main reasons for accepting ≤50% of the meals were described. Descriptive statistics for sample characterization and Kappa test agreement were calculated. Results: The same answer regarding the acceptance percentage was observed in 78.3% of cases for breakfast and 75.0% for lunch. The main reasons for accepting ≤50% were lack of appetite at breakfast (51.7%) and lunch (36.1%). The Kappa coefficient agreement between patients and evaluator reports on leftovers was k=0.696 and 0.692 for breakfast and lunch, respectively. Conclusions: Due to considerable agreement, the tool is validated for use with hospitalized adults and elderly in hospitals with the same profile as the institution studied and contributed to elaborate more effective dietary plans specific to the patient’s disease and nutritional status.