BACKGROUND: Latinxs are vastly underrepresented in mental health research; one of many contributing factors may be complexities in the research consent process, including language preferences. We examined determinants of comprehension of research consent procedures and tested the effects of a preconsent research schema condition among 180 adults with schizophrenia (60 Latinx-English and 60 Latinx-Spanish preference, and 60 non-Latinx White). STUDY DESIGN: Participants were randomly assigned (equal allocation) to an educational session regarding clinical research concepts and processes (schema condition) or to an attention control. Following a subsequent simulated consent procedure for a hypothetical drug trail, comprehension of consent disclosures was measured with 2 standard measures. STUDY RESULTS: One-way ANOVAs showed significant medium effect size differences between ethnicity/language groups on both measures of comprehension (η2s = 0.066-0.070). The Latinx-Spanish group showed lower comprehension than non-Latinx White participants; differences between the 2 Latinx groups did not reach statistical significance. Group differences were not statistically significant after adjusting for differences in education, or on scores from structured measures of acculturation, health literacy, or research literacy. Two-way ANOVAs showed no significant main effects for consent procedure on either comprehension measure (Ps > .369; partial η2s < 0.006) and no significant group-by-consent interactions (Ps > .554; partial η2s < 0.008). CONCLUSIONS: Although the preconsent procedure was not effective, the results suggest health and research literacy may be targets for reducing disparities in consent comprehension. The onus is on researchers to improve communication of consent information as an important step to addressing health care disparities.