Little is known about relations between maternal prenatal stress and specific cognitive processes-learning and memory-in infants. A modified crib-mobile task was employed in a longitudinal design to test relations between maternal prenatal cortisol, prenatal subjective stress and anxiety, psychosocial variables, and learning and memory in 3- and 5-month-old infants. Results revealed that maternal prenatal cortisol was affected by particular psychosocial variables (e.g., maternal age, whether or not the infant's grandmother provided childcare, financial status), but was unrelated to measures of maternal depression, anxiety, and stress. Although maternal prenatal cortisol was not predictive of learning or memory performance in 5-month-old infants, higher levels of basal maternal cortisol and reduced prenatal cortisol response was predictive of some learning and short-term memory measures in 3-month-old infants. These results suggest an influence of maternal neuroendocrine functioning on fetal neurological development, and the importance of separate examination of subjective and biological measures of stress.