We offer a neurobiologic theory based on animal work that helps account for the conspicuous male predominance in autism spectrum disorders (ASD). In young male animals, testosterone (TST) binds to androgen receptors (AR) in brainstem neurons responsible for enhancing brain arousal. As a consequence, arousal-related neurotransmitters bombard the amygdala hypersensitized by TST acting though AR. Arousal-related inputs are known to prime amygdaloid mechanisms for fear and anxiety, with resultant social avoidance. We hypothesize that similar mechanisms contribute to autism's male predominance and to its defining impaired social skills. The theory rests on two key interacting factors: the molecular effects of TST in genetically vulnerable boys in combination with environmental stresses they experienced in utero, neonatally, or during the first years. We postulate that higher TST levels and, therefore, higher amounts of arousal-related inputs to the amygdala sensitize these genetically vulnerable male infants to very early stresses. In sharp contrast to boys, girls not only do not have high levels of TST-facilitated arousal-causing inputs to the amygdala but they also enjoy the protection afforded by estrogenic hormones, oxytocin, and the oxytocin receptor. This theory suggests that novel technologies applied to the molecular endocrinology of TST's actions through AR will offer new avenues of enquiry into ASD. Since the high male preponderance in autism is important yet understudied, we offer our theory, which is based on detailed neurobehavioral research with animals, to stimulate basic and clinical research in animals and humans and hopefully help develop novel more effective medical treatments for autism. Autism Res2011,4:163–176. © 2011 International Society for Autism Research, Wiley Periodicals, Inc.