Azidothymidine (AZT) is administered to pregnant women with HIV to prevent the spread of infection to their fetuses. Since gestation is a period of critical neurodevelopment, it is important to determine the risk AZT exposure may pose to neurobehavioral function of the offspring. The current study focused on teratological risks of developmental AZT exposure to neurocognitive function. Male and female Swiss mice were administered AZT or vehicle (0, 100, or 200 mg/kg/day po given twice daily in equal amounts for 32 weeks before and during gestation). Adult male and female offspring (n = 10/sex/treatment group) underwent neurobehavioral testing focused on determining learning and memory capabilities in the radial-arm maze. AZT exposure did not cause significant deficits during radial-arm maze acquisition. No impairment was seen in asymptotic levels of choice accuracy indicative of working memory function. Attempts to unmask subtle learning impairments following developmental AZT by the introduction of behavioral challenges such as reduction of motivational state (food restriction either 4-6 h or 22-24 h) or imposition of intrasession delays of 1.5 min to 2.5 h were unsuccessful. With a 4-week intersession delay, a significant AZT Treatment x Delay effect was seen with a significantly greater decline seen in the controls as compared to the 100 mg/kg/day AZT group. Locomotor activity on the radial-arm maze was significantly affected by AZT treatment (100 mg/kg/day) during the acquisition phase, but not during the other test phases. No behavioral alterations were seen related to stress as measured by the elevated plus maze. Vestibulomotor functioning on the balance beam remained unaltered. Using an extended dosing regimen including dosing of both sires and dams, as well as placing a greater demand on reproductive system performance with three continuous breedings, this study detected only subtle neurobehavioral impairments in mice after prenatal AZT exposure at clinically relevant doses.