Code-switching is a widespread phenomenon among bilinguals and a consistent part of their linguistic behavior across lifespan (de Bruin, Samuel, & Duñabeitia, 2020) and socioeconomic status (Casas, 2016; Poplack, 1980). As such, understanding the cognitive aspects of code-switching becomes an important part of the study of bilinguality. However, research on the processing of code-switching is still scarce. Here we investigate properties of verbal morphology which may modulate bilingual sentence processing. Much evidence in the bilingual literature indicates that Spanish-English bilinguals more easily process code-switches in progressive constructions (e.g., Los niños están playing at the park, ‘The children are…’) than in perfect constructions (e.g., Los niños han played at the park, ‘The children have…’) (Guzzardo-Tamargo, Valdés Kroff, & Dussias, 2016). Furthermore, evidence from highly proficient L2 learners of English (Pliatsikas & Marinis, 2013) shows that regular and irregular past tense verbs are processed differentially, with shorter RTs for irregular than for regular verbs. We investigate the role of verbal morphology in Spanish-to-English code-switching using past participles, assuming that they are conceptually the same as past tense verbs (Halle & Marantz, 1993). Our study uses a self-paced reading (SPR) paradigm to investigate how bilinguals (N = 33; MAge = 24.5, SD = 5.2, range 18-38) living in the U.S. process progressive/perfective code-switching structures, further manipulating i) regularity of the participle form (regular vs. irregular), and ii) verb class (Latinate vs. non-Latinate) (for an example, see Table 1). In line with previous findings, we expect to see a main effect of aspect (progressive/perfect), but it remains an open question whether the differential processing of regular vs. irregular verbs reported in Pliatsikas & Marinis (2013) for L2 learners holds in bilingual code-switching and whether differences in verb class result in differential processing. Word reading times (RT) for the participle and the word that followed it (the spillover) were inverse-transformed to −1000/RT, and extreme values (> 2.5 mean absolute deviations) were trimmed. To adjust for differences in the numbers of characters in the progressive and perfect forms, word length in characters was included as a predictor in the models. A linear mixed effects model for RT at the participle showed no significant effects of aspect or regularity, and no significant interaction. However, for the non-Latinate verb class, a linear mixed effects model of RT at the spillover showed a significantly shorter RT when the verb was irregular and progressive than when it was irregular and perfect (b = −0.13, t = −1.97, p < .05; see Figure 1). These results suggest that code-switching within verbal phrases is affected by aspect in irregular non-Latinate verbs as reflected in RTs in the spillover region. We will discuss potential implications of our findings. Email: dcastillo1@gradcenter.cuny.edu<mailto:dcastillo@gradcenter.cuny.edu>