Despite the popularity of neutralization theory, numerous issues about the theory still exist. This study examines the stability of neutralization acceptance, the role of neutralization acceptance on self-reported delinquency, and the impact of social bonds on neutralization acceptance. Analyses were conducted using data culled from one cohort of respondents to the Denver Youth Survey. These issues are explored via structural equation modeling with measures of neutralization acceptance that are contextualized to specific subtypes of delinquency. The findings suggest that neutralization acceptance is a more robust predictor of delinquency than delinquency is of neutralizing attitudes, and that some of these effects are generally stable across time. However, many of these effects operate differently across types of delinquency and sometimes across levels of social bonds. Implications for theoretical development and for future research are discussed. Abstract published by arrangement with Sage Journals.