The success rate of captive elephant breeding programs worldwide is poor. Along with undiagnosed reproductive disorders in females and fatal diseases such as the newly discovered herpesvirus infection, male infertility now is considered a major contributing factor in the failure to maintain self-sustaining captive populations. To address questions related to male reproductive dysfunction, approximately 309 ultrasonographic assessments combined with semen collection were performed in captive (n = 10) and wild (n = 4) African (Loxodonta africana) and captive (n = 61) Asian (Elephas maximus) elephants. Bulls ranged from 4 to 50 years of age and were examined at 9 institutions in North America, 13 in Europe, 2 in Africa, and 7 in Asia. About half of the reproductive assessments were performed in protected contact situations with elephants handled in a restraint device, and half involved assessments of trained Asian bulls managed in free contact. Four wild African and two Asian elephant bulls were evaluated after receiving general anesthesia. Transrectal ultrasound was used to characterize the morphology and functionality of the entire urogenital tract, including the testes and accessory sex organs. Bulls were categorized on the basis of breeding status (breeders vs. non-breeders) and social history (i.e., type of interaction with conspecifics and keepers). Most of the bulls were non-breeders (designated Types I–V). Type I (n = 3 African, 6 Asian) and Type V (n = 1 Asian) were immature and castrate, respectively. On the basis of keeper evaluations, Type II bulls (n = 2, 4) were subordinate to older cows and keepers, whereas Type III bulls (n = 4, 28) were dominated by other bulls. Type IV (n = 1, 8) were older bulls of unknown history that exhibited numerous testicular pathologies resulting in poor semen quality. Non-breeding bulls included those that were exposed to females, but failed to breed, as well as those that had no opportunities to breed. Type VI individuals (n = 4, 14) were proven breeders. The percentage of observable reproductive tract pathology in adult males was remarkably low (14%), even in older bulls. However, apparent infertility of non-organic cause (i.e., not due to specific anatomical abnormalities) in these otherwise healthy bulls was high (32%). Semen quality varied markedly in ejaculates collected from the same bull, as well as from different bulls. In conclusion, although many of these bulls could serve as semen donors for natural mating or artificial insemination, the inconsistent production of good-quality ejaculates raises questions as to the reliability of these individuals to participate in breeding programs. The apparent inhibitory effect of suppressive social interactions on reproductive potential also needs to be investigated. Ultrasound examinations combined with semen collection should be conducted periodically to estimate the reproductive value of each bull and determine whether altered management strategies are needed to enhance captive breeding. Zoo Biol 19:333–345, 2000. © 2000 Wiley-Liss, Inc.