To investigate the efficacy and efficiency of environmental taxes, I analyze the rate at which taxes on motor gasoline are passed to consumers by estimating two cointegrating vector autoregression models for each of six states. For state models that specify the retail price of motor gasoline without taxes, exclusion tests suggest that taxes on motor gasoline are not passed to consumers on a one-for-one basis. For state models that specify the retail price of motor gasoline including taxes, results indicate that taxes are passed to wholesale prices in Florida and Massachusetts on a one-for-one basis and are passed to retail prices with a 'mark-up' in Florida, Massachusetts, New York, and Ohio, and are not fully passed through in Washington. State-specific rates of pass through differ from results suggested by theory and fixed effects estimators, which may be biased by the presence of nonstationary data and the assumption that the rate of pass through is the same across states. Rates of pass through greater than one transfer $12.2 billion from consumers to retailers in FL, $2.3 billion in MA, and $19.2 billion in NY during the sample period, which represent 10.7%, 6.0%, and 23.9% of total expenditures on regular motor gasoline.