We have used combinatorial algebra and computer simulations to calculate expected means, variances, and probabilities of hormone-peak coincidences in 2 or more endocrine pulse series. We illustrate application of this conditional probability analysis to pulsatile LH, FSH, and/or PRL data. We observed that 1) serum LH and FSH pulses in 14 young men were randomly associated on different days (P greater than 0.10), but highly synchronized on any given day (P less than 0.0001); 2) serum LH and FSH (P less than 0.0001), LH and PRL (P = 0.023), and FSH and PRL (P = 0.0003) peaks were significantly coupled in healthy postmenopausal women; and 3) the number of triple coincidences among LH, FSH, and PRL release episodes in postmenopausal women significantly exceeded chance expectations (P less than 0.0001). We conclude that suitable statistical coincidence analysis can offer an informative tool with which to evaluate nonrandom event concordance in endocrine investigations, such as clinical studies of the temporally coordinated release of anterior pituitary hormones.