OBJECTIVE Malnutrition is emerging as a multidimensional concern of ageing with a high prevalence among nursing home residents. This study investigated the extent of malnutrition among old subjects in family practice and its relationship with major complications. METHODS Over 75 years old subjects (n = 274) filled the Mini Nutritional Assessment questionnaire. Appearance of major events in the following 6 months were registered. RESULTS MNA scored were 11.5 ± 3.1, with 175 (64%) subjects showing no malnutrition, 69 (25%) resulted at risk, and 30 (11%) malnourished. Within at risk group, 1.4% was resident, 7% bed rested, 8% had a history of major bone fracture, 33% was demented and 24.6% hospitalized at least once in the last year. Among malnourished patients, 10% was resident, ten bed rested with 70% showing multiple bedsores, 20% have had bone fractures, 60% were demented and 13% hospital admitted in the previous year. In over 90% of them, malnutrition had neither diagnosed nor considered before. During follow-up, a significantly higher number of major events including death occurred in the malnourished group. By multivariate logistic regression, n = 56 (20.4%) patients resulted at risk of major complications. The sensitivity of the questionnaire in identifying these patients was 84% with the cut-off value of 7 associated with the highest prediction (positive predictive value, 0.92; negative predictive value, 0.71) yielding a specificity of 92%. CONCLUSIONS The prevalence of malnutrition is high among older subjects in the setting of family practice. The Mini Nutritional Assessment allows to identify malnourished subjects better than BMI and effectively predicts the risk of major events.