OBJECTIVES: To identify predictors and consequences of nutritional risk, as determined by the Mini Nutritional Assessment (MNA), in older recipients of domiciliary care services living at home. DESIGN: Baseline analysis of subject characteristics with low MNA scores (<24) and follow‐up of the consequences of these low scores. SETTING: South Australia. PARTICIPANTS: Two hundred fifty domiciliary care clients (aged 67–99, 173 women). MEASUREMENTS: Baseline history and nutritional status were determined. Information about hospitalization was obtained at follow‐up 12 months later. INTERVENTION: Letters suggesting nutritional intervention were sent to general practitioners of subjects not well nourished. RESULTS: At baseline, 56.8% were well nourished, 38.4% were at risk of malnutrition, and 4.8% were malnourished (43.2% not well nourished). Independent predictors of low MNA scores (<24) were living alone, and the physical and mental component scales of the 36‐item Short Form Health Survey. Follow‐up information was obtained for 240 subjects (96%). In the ensuing year not well‐nourished subjects were more likely than well‐nourished subjects to have been admitted to the hospital (risk ratio (RR) = 1.51, 95% confidence interval (CI) = 1.07–2.14), have two or more emergency hospital admissions (RR = 2.96, 95% CI = 1.15–7.59), spend more than 4 weeks in the hospital (RR = 3.22, 95% CI = 1.29–8.07), fall (RR = 1.65, 95% CI = 1.13–2.41), and report weight loss (RR = 2.63, 95% CI = 1.67–4.15). CONCLUSION: The MNA identified a large number of subjects with impaired nutrition who did significantly worse than well‐nourished subjects during the following year. Studies are needed to determine whether nutritional or other interventions in people with low MNA scores can improve clinical outcomes.