The statistical power of a t-test was evaluated comparing exposed and control groups for hormone concentration values of single blood samples versus three samples drawn 20 minutes apart. The decrease in statistical power due to an estimated decrease in participation in response to multiple blood sampling was also studied. Blood samples were drawn from 45 men, who provided monthly semen samples over a 9 month study period, and analyzed for individual endocrine profiles. Three blood samples were drawn 20 minutes apart, on three occasions, between 8AM and 8PM, during the study which included June, October, and February. Determinations were made of serum concentrations of follicle stimulating hormone, luteinizing hormone, testosterone, and prolactin. Variability was estimated between the 20 minute blood draws using a component of variance model. The three blood draws provided a marginal improvement over a single blood draw in detecting population shifts. The authors conclude there should be better compliance when only a single sample is required and participation should be improved with the single draw.