Fractured neck of femur occurs mostly in the older female population and is generally caused by falls. Malnutrition has been postulated as a factor that increases the tendency to suffer falls. Nutritional status of older female hospital patients admitted for emergency surgery for fractured neck of femur recruited (n 75), was compared with an age-matched independent-living group of females attending one of three local day centres (n 50). Dietary assessment was undertaken using three consecutive 24 h dietary recalls and, in the hip fracture group, completed menu cards were used as memory prompts. Data concerning key lifestyle characteristics were obtained using a face-to-face administered questionnaire. Blood samples were taken to determine levels of plasma albumin, transferrin, C-reactive protein (CRP), cholesterol, vitamin C, Se, Zn and total antioxidant status. Haemolysate samples were analysed for Se-dependent glutathione peroxidase activity. There were no significant differences in age between the two groups, but the hip fracture patients had lower mean values for body weight (59.6 v. 67.5 kg; p = 0.005), mindex (weight/demispan) (83· v. 94.4 kg/m; P < 0·0001), calculated BMI (24·1 v. 27·5 kg/m2P<0·0001), mid-upper arm circumference; 27·1 v. 31·3 cm, p = 0.001) and triceps skinfold thickness; 17·0 v. 18.9 mm, p = 0·005) than the home-living group. The hip fracture patients had lower intakes of energy (4·3 v. 5·4 MJ, p = 0·001), fat (p = 0·025), carbohydrate (p = 0·002), protein (p = 0·006), thiamine (p = 0·017), vitamin B6 (p = 0·001), calcium (p = 0·01), K (p = 0·001), Mg (p = 0·001), P (p = 0·001), Fe (p = 0·007), Se (p = 0·008) and NSP (p = 0·001). Mean intakes of both groups were below the estimated average requirement for energy and below the reference nutrient intakes for folate, Ca, vitamin D, Mg, K, Se and Zn. In a high percentage of the hip fracture group the dietary intake of particular nutrients fell below the lower reference nutrient intake for Se (73 %), Mg (54 %) and Fe (19 %). As expected, the fracture patients had reduced plasma albumin (P<0·0001) and increased CRP (P<0·001) values. They had higher plasma vitamin C levels (P<0·001) and lower cholesterol levels (p = 0.04) than the day centre attendees. There were no significant differences in plasma levels of Se, Zn, transferrin or haemolysate glutathione peroxidase activity between the two groups. However, there was evidence of under-nutrition in both groups as key anthropometric values were low, plasma nutrient and metabolite levels were below the standard reference ranges and many individuals had low dietary intakes for specified nutrients.