According to modernization theory, enforcement schemes that rely on end-of-the-pipe regulation are not as effective at achieving improved environmental performance as market-based approaches that encourage pollution prevention. Consistent with that observation, the U.S. Environmental Protection Agency transitioned to the use of self-policing to encourage pollution prevention. Other studies note that environmental compliance is significantly affected by traditional “command-and-control” strategies. Using Prais Winston regression we examine these contrasting views by estimating the relationship between toxic releases, self-policing, and criminal prosecutions from 1988 through 2014. Initial correlations suggest that (1) self-policing is not associated with toxic releases but that (2) criminal prosecutions may reduce toxic releases through general deterrence signals. Subsequent analyses controlling for gross domestic product revealed that neither self-policing nor criminal enforcement correlate with toxic releases but that gross domestic product was the strongest predictor of emissions. The implications of these findings for the control of toxic emissions are discussed.