A collective critical review of the literature on postsplenectomy sepsis from 1952 to 1987 has been undertaken. The reports cover a cohort of 12 514 patients undergoing splenectomy but of these only 5902 reports were sufficiently detailed to allow a useful analysis. The incidence of infection after splenectomy in children under 16 years old was 4.4 per cent with a mortality rate of 2.2 per cent. The corresponding figures for adults were 0.9 per cent and 0.8 per cent respectively. The present analysis of well documented patients has shown that severe infection after splenectomy for benign disease is very uncommon except in infants (infection rate 15.7 per cent) and children below the age of 5 years (infection rate 10.4 per cent). Many of these reported postsplenectomy infections may have been coincidental. It is also apparent that children contract a different type of infection after splenectomy than adults, predominantly a meningitis which is less frequently fatal. Adults, in contrast, appear to develop a septicaemic type of illness associated with a higher mortality rate. This survey has also shown that children are reported to be no more susceptible to pneumococcal sepsis than to infection caused by any other organism. Although the removal of the spleen in otherwise normal people does not appear to be associated with an increased frequency of infection, the presence of a coexistent disorder, notably hepatic disease, can increase the risk substantially.