Insensitivity of tests for reproductive problems.

Studies of reproductive capacity in rodents have been conducted for decades using a design which has had little refinement over the years. Most commonly, one or two litters from each of one through three generations are raised. The variability between studies, generations, groups, and animals often precludes the resolution of subtle toxicologic effects on reproductive parameters. Nonreproductive parameters yield more reliable data than reproductive parameters. Changes in design and additions to the protocol which will provide additional data are discussed. These include semen evaluation and the use of larger group sizes in reproduction studies.