This paper deals with the perception of nonnative vowel contrasts by highly proficient bilinguals, as well as the relationship between these perceptual processes and spoken word recognition processes in a bilingual’s nonnative language. To this end, in section I, I introduce some fundamental issues in L2 production and perception, briefly review select theoretical models of L1 constraints on the perception and acquisition of nonnative contrasts by naïve and L2 listeners, and review some empirical evidence that bears on the issue of plasticity. In section II, I discuss the potential implications of inaccurate perception of nonnative contrasts on word recognition processes in the L2. In particular, I review a study by Pallier et al (2001) and suggest an alternative account of the findings. In section III, I present two experiments that aim to replicate and extend the findings of Pallier et al (2001) in a group of Spanish-English bilinguals. In particular, we asked whether the availability of phonological features in the native language grammar predicts the acquirability of nonnative vowel contrasts. We found that Spanish learners of English performed poorer on contrasts for which their native language contains the necessary phonological feature, contra the predictions of the feature redeployment hypothesis (Brown, 1998; 2000). Finally, in section IV, I discuss the implications and limitations of the experiments, as well as potential avenues for future research.