Objective: To analyze to role of the Short-Form Mini Nutritional Assessment in the nutritional assessment of elderly individuals with cancer. Study Design: Cross-sectional epidemiological study. Location: Outpatient services at eight cancer hospitals in the cities of Campo Grande and Rio de Janeiro, in Brazil. Participants: 333 individuals, aged 60 or older, with prostate cancer, interviewed in the period before the definition of cancer treatment. Methods: The nutritional profile was assessed with the Short Form Mini Nutritional Assessment (MNA-SF) and with the Body Mass Index (BMI), according to Lipschitz’s classification. Socio-demographic and clinical variables were also considered. We analyzed the distribution of elderly individuals by MNA-SF items according to MNA-SF and BMI categories and the Fisher’s exact test was applied for testing the statistical significance of the observed differences (p≤0.05). Results: Based on the MNA-SF, 235 (73.0%) individuals had an adequate nutritional status; 74 (23.0%) were at risk for malnutrition and 13 (4.0%) were considered malnourished. Based on the BMI, 49 elders (15.2%) were classified as low weight; 150 (46.4%) as normal weight and 124 (38.4%) as overweight. For each item in the MNS-SF, the frequency of answers corresponding to deficiencies was high among elderly individuals at risk for malnutrition, with statistical significance. With respect to the BMI, only “Food intake declined over the past 3 months” and “BMI” had statistically significant differences. Conclusion: MNA-SF showed a good relationship between its component items and proved to be an adequate tool to describe nutritional risk in elderly cancer patients. Early detection of this risk paves the way for an early nutritional approach and for preventing undesirable outcomes with respect to the health of those individuals. As other advantages, MNA-SF is easy to interpret, demands a short time for application and is well accepted by the elders.