A simple generalized linear model of toxoplasmosis incidence has been applied to serologic data from more than 9,000 women who delivered babies in Switzerland. This model, based on an assumption of constant incidence of toxoplasmosis throughout time and ages, yields estimates of incidence rates that show marked contrasts between subgroups of women categorized on the basis of their geographic origin and duration of residence in Switzerland. The patterns observed are in agreement with previously reported data from specific areas. When applied to Swiss resident women of different age groups, the average incidence rate estimates obtained were remarkably similar. This was not the case among the subgroup of Portuguese immigrant women, nor when the model was applied to previously published data from Sweden where toxoplasmosis incidence has clearly decreased in the last 20 years. Although models for the analysis of cross-sectional data involve major simplifying assumptions, it appears that they can yield useful epidemiologic information and that departures from the assumed simplified structure of the data can sometimes be identified.