The Ontario Farm Family Health Study was designed to assess retrospectively the potential adverse effects of exposure to pesticides on pregnancy. Information on the health and life style of approximately 2,000 farm couples, as well as a history of use of pesticides on the farm, was collected by questionnaire. This analysis focuses on pre- and postconception exposure to phenoxy herbicides and the risk of spontaneous abortion using the complete (to date) pregnancy history for each woman. Preconception exposure (from 3 months before conception to the month of conception) was weakly associated with the risk of spontaneous abortion at <20 weeks' gestation [adjusted odds ratio (OR) = 1.1; 95% confidence interval (CI) = 0.6-1.9]. When the analyses were restricted to spontaneous abortions of <12 weeks, the risk was more than doubled (adjusted OR = 2.5; 95% CI = 1.0-6.4), but the results were sensitive to the cutpoint used. If the husband did not normally wear protective equipment during application, the crude OR for early spontaneous abortions was 5.0 (95% CI = 0.7-36.2). Exposure to phenoxy herbicides during the first trimester was generally not associated with increased risk of spontaneous abortion. The results suggest a possible role of preconception (possibly paternal) exposures to phenoxy herbicides in the risk of early spontaneous abortions.