Study data came from the Youth in Transition survey conducted in 1966 by the Institute for Social Research, University of Michigan. A nationwide representative sample of 2,213 10th-grade boys answered questions about the consistency and method of their discipline, as well as possible delinquent behavior. Discipline had several dimensions: physical punishment, defined as slapping, was either consistent or intermittent in its application, and parental demands were either consistent or inconsistent. Consistent demands meant that parents made their expectations clear, whereas inconsistent demands meant that parents did not make expectations clear and children were confused. Physical punishment increased delinquency only in the presence of inconsistent demands. Intermittent use of physical punishment had the same effect as consistent use. In contrast, most previous studies found a positive relationship between physical punishment and delinquency because they failed to control for the consistency of parental demands. Physical punishment is not recommended because it promotes delinquency unless parental demands are unusually consistent, its deterrent effect is very slight even with consistent demands, and it may have adverse consequences such as increased neuroticism in the child. Two data tables, 21 references, and an appendix presenting the self-report scale called 'Seriousness of Delinquency' are included.