Regression analysis of the relationship of serum circulating anodic and cathodic antigens (CAA and CCA), as a possible direct measure of worm burden, and fecal egg counts allows the study of phenomena like density-dependent fecundity in human Schistosoma mansoni infections. For a reliable analysis, variations in egg count measurements as well as in circulating antigen levels have to be taken into account, and an accurate estimation of these variations (represented by parameter λ in the so-called Deming regression) is of great importance. From a new, extensive data set of repeated measurements of fecal egg counts and CAA and CCA concentrations we determined the respective values for parameter λ, and (re)analysed the relationship between circulating antigens and egg counts in 3 data sets from Burundi, Senegal and Zaire by Deming regression. For comparison, ordinary linear regression was performed as well, which considerably biased the regression lines for CCA, but not for CAA. The analyses resulted in a clearly non-proportional relationship between egg counts and CAA, and, to a lesser extent, CCA. Assuming that egg counts and antigen measurements directly reflect egg production and worm burdens, respectively, our findings reinforce the indication of density-dependent fecundity in schistosomiasis mansoni, as suggested by others.