Shortage of donor livers has led several liver transplant centres to widen their definition of liver donor suitability. We have assessed the function of liver grafts from "marginal" donors and attitudes to use of such organs. Over an 18-month period, livers used in 30 of 213 consecutive liver transplantations in Birmingham, UK, came from marginal donors (history of alcoholism, abnormal liver function test results, drug overdose that included paracetamol, advanced cardiovascular disease, sepsis, lengthy hypotension [systolic blood pressure < 80 mm Hg for > 1 h], high-dose inotropic drug use). 16 of these donors had been refused by other UK liver transplant centres, 11 on medical grounds. The controls were grafts retrieved from "good" donors (n = 183) during the same period. All 30 grafts showed satisfactory early function but had greater day 1 (p = 0.004) and peak serum aspartate aminotransferase (p = 0.0008) values than control grafts. Graft and patient survival at 1 year in the two groups was similar (72% vs 73% and 80% vs 82%, respectively). To assess attitudes to marginal donor livers, a questionnaire outlining the details of these 30 donors was sent to the 80 centres in the European Liver Transplant Group, and 60 replied. Median immediate refusal rate of the marginal donors was 7/30 (range 0-18) and median outright acceptance rate was only 11/30 (1-26). Larger centres were less selective, with a significantly lower refusal rate (p = 0.03). These results indicate that, because of existing liver donor criteria within Europe, usable donor livers are being unnecessarily refused on medical grounds.