Background. The extension of living donor liver transplantation to adult recipients is limited by the adequacy of the size of the graft. We evaluate the effect of the graft size on the survival of the recipient in order to establish a clinical guide for the minimum requirement. Methods. The clinical records of 14 adults and 11 children (body weight 6.1-100 kg) who underwent living donor liver transplantation for chronic or acute liver failure were reviewed. The effect of the graft weight ratio (graft weight divided by standard liver weight of recipient) on graft function and survival was studied. Results. The graft weight ratio ranged from 31 to 203%. The overall graft and patient survival rates were 84% at a median follow-up of 29 months. The survival rate was 95% for recipients with a graft weight ratio >40%, and 40% only for those with a ratio ≤40% (P=0.016). It was 88% (7/8) when the ratio was >100%, 100% (5/5) when the ratio was 71 to 100%, 100% (7/7) when the ratio was 41 to 70%, and 40% (2/5) only when the ratio was ≤40%. When the graft weight ratio was ≤40%, early graft dysfunction was evident and contributed to the causes of death in three patients. Conclusions. Preoperative computed tomographic measurement of liver size of a living donor is essential. A graft that represented 40% or less of the recipient's standard liver weight should be regarded as a marginal graft with a lower success rate.