NIOSH and the University of Minnesota conducted a reproductive health study of men applying pesticides in northwest Minnesota. Two semen samples were collected from each of 90 study participants. The first semen sample was collected in July 1998 at the end of the herbicide application season and the second in late October 1998 at the end of the fungicide application season. A complete semen analysis was conducted on each semen sample including computerized semen analysis (HTM-lVOS, Hamilton-Thorn, Beverly, MA). The semen data were stratified by pesticide application history and analyzed for the effects of pesticides using generalized estimating equations (for discrete outcome variables) and the mixed model repeated measures approach (for continuous outcome variables). The percent normal sperm morphology(WHO) was significantly lower in the fungicide applicators (non-fungicide 29.8 % +/- 2.6; fungicide 20.8% +/- 3.4; P = 0.016). Fungicide applicators also had a lower sperm straight line velocity than the men not applying fungicides (non-fungicide 55.7 microm/sec +/- 1.3; fungicide 48.4 microm +/- 2.6; p = 0.02). A decrease in straightness of swimming path (VSLNCL) was associated with both herbicide and fungicide application (non-herbicide 0.60 +/- 0.03, herbicide 0.52 +/- 0.01, P = 0.02; non-fungicide 0.58 +/- 0.01, fungicide 0.54+/- 0.02, P = 0.05). The associations found in these analyses generate the need to separate the broad categories of herbicides and fungicides and study the association between specific pesticide chemicals and human semen quality. Further data analyses are underway detailing specific pesticide use based on current pesticide application records.