Maximal voluntary grip strength has been measured in normal adults aged 18-70 years (17 f, 18 m) and compared with other indices of body muscle mass. Grip strength (dominant side) was directly proportional to creatinine excretion (r = 0.81); to forearm muscle area (r = 0.73); to upper arm muscle area (r = 0.71) and to lean body mass (r = 0.65). Grip strength relative to forearm muscle area decreased with age. The study of a subgroup of normal subjects revealed a small but significant postural and circadian effect on grip strength. The effect on maximal voluntary grip strength of sedatives in elderly subjects undergoing routine endoscopy (n = 6), and of acute infections in otherwise healthy individuals (n = 6), severe illness in patients requiring intensive care (n = 6), chronic renal failure (n = 7) and anorexia nervosa (n = 6) has been assessed. Intravenous diazepam and buscopan produced a 50 per cent reduction in grip strength which returned to normal within the next 2-3 h. Acute infections reduced grip strength by a mean of 35 per cent and severe illness in patients in intensive care by 60 per cent. In patients with chronic renal failure grip strength was 80-85 per cent of that predicted from forearm 'muscle area' (P less than 0.05). In anorectic patients the values were appropriate for their forearm muscle area. Nevertheless nutritional rehabilitation of one anorectic patient did not lead to a consistent improvement in grip strength.