The use of immunodiagnostic techniques for studying epidemiological aspects of schistosomiasis have been of limited value in the past owing to the type of techniques adopted and the vide cross reactivity exhibited by antigens used in the tests. In this study a relatively new technique, the Enzyme Linked Immunosorbent Assay (ELISA) was used. After the development of a standardised procedure it was applied to a wide range of S. mansoni infections representative of different epidemiological situations. A number of modifications in the basic enzyme assay technique were introduced and their merits discussed. Those which proved to be of value were a) the use of ultracentrifuged antigens, b) the development of a method for the determination of the optimum value for the reference serum endpoint, c) the use of serum eluted from filter paper blood spots, d) the use of precoated antigen plates prepared by air or vacuum drying and e) the development of quality control procedures. In evaluating the diagnostic performance in human schistosomiasis ELISA was compared with conventional immunodiagnostic techniques such as skin tests, complement fixation tests (CFT), indirect fluorescent antibody tests (IEAT) and a recently developed radioimmunoassay (RIA)incorporating a species specific egg antigen. Sera from E. Africa and the West Indies were used to assess the test in relation to important diagnostic criteria. The ELISA using soluble egg antigen exhibited higher levels of sensitivity and specificity compared to skin teats, CIT and IPAT with sera from schistosome endemic and controls from non-endemic areas. Unlike the RIA however cross reactions occurred with heterologous schistosome infections and the test did not show the same degree of association with intensity of infection and responsiveness to chemotherapy. In two large control programmes, in the Gezira region of the Sudan where S. amsoni is hyperendemic, and St. Lucia where prevalence is low, an evaluation of the diagnostic reliability of the test and the relevance of antibody measurements was made in two widely different epidemiological situations. The relative merits of a serological as compared with a conventional parasitological approach in epidemiology were examined with particular emphasis on the use of serodiagnosis for measuring incidence rates and as a primary screen for the selective removal of uninfected individuals in a prevalence survey and prior to chemotherapy.