This paper presents a method of estimating the potential global morbidity due to human intestinal nematode infections (Ascaris lumbricoides, Trichuris trichiura and hookworms), based on the observed prevalence of infection. The method relies on the observed relationships between prevalence and intensity of infection, and between worm burden and potential morbidity. This approach is shown to be sensitive to the precision of the original prevalence estimates and, in particular, to the degree of spatial heterogeneity in levels of infection. The estimates presented here indicate that several tens of millions of children are likely to suffer developmental consequences from infection, and suggest that the global disease burden of geohelminthiasis may be significantly greater than was supposed.