When test article-related testicular toxicity or Leydig cell tumors are identified in nonclinical studies, the measurement of circulating hormones such as luteinizing hormone, follicle-stimulating hormone, inhibin, testosterone, or prolactin is often considered in order to aid mechanistic investigations or to identify potential biomarkers in man. Although some hormone levels are relatively constant, others are subject to wide variability owing to pulsatility of secretion, diurnal rhythms, and stress. To avoid being misled, it is important that this variation is factored into any study design that includes hormone measurements. Since all these possibilities start from the pathologist's reading of the tissue sections, we begin with a review of the morphologic changes that are tied to underlying alterations in hormones. We then provide the reader with basic information and representative hormone data, including coefficients of variation, for the major male reproductive hormones in the three main nonclinical species (rats, dogs, and cynomolgus monkeys). Power and probability tables for rats and dogs allow estimates of the number of animals or samples needed to provide a given likelihood of detecting a hormonal change of a given size. More importantly, we highlight the variability of this process and the real value in readers developing this information at their own site.