There has been a recent resurgence of interest in the use of human albumin solution (HAS) for the resuscitation of critically ill patients. In particular, it may be beneficial for patients with severe sepsis and septic shock. We outline the evidence for the biological plausibility of this effect, perform a meta-analysis of randomised controlled trials comparing the effect of HAS with other fluid resuscitation on mortality in this group of patients and discuss the likely contribution of recently completed trials in this area. We included nine trials of 1,435 patients in the meta-analysis. Although HAS resuscitation was associated with a trend to lower mortality (relative risk 0.90, 95% confidence interval 0.79–1.02), we found that most trials reported to date are small and of variable methodological quality. The questions of a potential differential effect of dose on outcome and the pharmacological variability of HAS products have not been adequately addressed to date. We conclude that the routine administration of HAS to patients with severe sepsis and septic shock is difficult to justify on the basis of current knowledge.