This study examines whether maternal warmth moderates the association between maternal use of spanking and increased child aggression between ages 1 and 5. Participants were 3,279 pairs of mothers and their children from a cohort study of urban families from 20 U.S. cities. Maternal spanking was assessed when the child was 1 year, 3 years, and 5 years of age. Maternal warmth and child aggressive behavior were measured at 3 years and 5 years of age. Models controlled for demographic characteristics (measured at the child's birth), child emotionality (measured at age 1), and maternal psychosocial risk factors (measured when children were 3 years old). Cross-lagged path models examined the within-time and longitudinal associations between spanking and child aggression. Results indicated that maternal spanking at age 1 was associated with higher levels of child aggression at age 3; similarly, maternal spanking at age 3 predicted increases in child aggression by age 5. Maternal warmth when children were 3 years old did not predict changes in child aggression between 3 and 5 years old. Furthermore, maternal warmth did not moderate the association between spanking and increased child aggression over time. Beginning as early as age 1, maternal spanking is predictive of child behavior problems, and maternal warmth does not counteract the negative consequences of the use of spanking.