In four experiments we studied the impact of retrieval-extinction training on the extinction and reinstatement of alcoholic beer seeking. Experiment 1 showed that preceding daily extinction sessions with a brief (10 min) extinction session (retrieval-extinction) attenuated the context-induced reinstatement of alcoholic beer seeking, thereby replicating and extending the findings of Xue et al. (2012). Experiment 2 then showed that the retrieval-extinction manipulation could attenuate the reinstatement produced by reversible inactivation of the nucleus accumbens shell prior to test. Experiment 3 showed that a modified extinction protocol that involved a reversed retrieval (i.e., extinction then retrieval) was also able to attenuate context-induced reinstatement. Finally, experiment 4 showed that the extinction-retrieval manipulation facilitated the reacquisition of alcoholic beer seeking as evidenced by increased breakpoints and responses during tests under a progressive ratio schedule. Taken together, these findings show that retrieval-extinction training protocols can alter the propensity to reinstate extinguished drug seeking but that these alterations are not always protective. These findings are inconsistent with accounts of the retrieval-extinction manipulation in terms of memory reconsolidation and deepened extinction. Instead, they are consistent with the notion that this manipulation increases the sensitivity of animals to the contingencies in effect during testing.