In a prospective study, we have examined the effect of nutritional status, using anthropometric measurements, on outcome in 102 consecutive adult patients undergoing elective orthotopic liver transplantation. Mid-arm muscle circumference was calculated from these two measurements. Patient outcome variables were time spent in the intensive therapy unit, total time in hospital, infective complications and mortality within 6 months. Graft outcome variables were early graft function, peak aspartate transaminase, alkaline phosphatase, bilirubin and prothrombin time. Group A patients were below and group B patients above the 25th percentile for mid-arm circumference and triceps skin fold thickness. Eighty-four patients (79%) were at or below the 25th percentile of anthropometric measurements and 30 patients (28%) were below the 5th percentile. The median mid-arm muscle circumference in group A was 22.3 (range 16.4-28.9) cm and 25.7 (range 21.7-31.8) cm in group B. The medial albumin level was similar in the two groups. There were significantly more bacterial infections in group A (27/84, 32%) than in group B (2/22, 8%; chi(2) = 5.4, P = 0.02). There was a difference in mortality up to 6 months post-operatively that failed to reach statistical significance (Wilcoxon-Gehan statistic -199, P = 0.09). There were 11/84 (13%) deaths in group A and no deaths in group B (chi(2) = 2.8, P = 0.09). Post transplantation, there were significant differences (Kruskal-Wallis Anova) between groups A and B for peak alkaline phosphatase (683 vs 334 IU/I, P = 0.05) and peak prothrombin time [16 (range 13-25) vs 19.5 (range 12-65), P = 0.03]. These data suggest that a significant proportion of patients undergoing liver transplantation are nutritionally compromised and that this has effects on patient infection, susceptibility, graft function and mortality, which may possibly be improved by nutritional intervention.