Objective: The exclusion criteria of tobacco cessation randomized clinical trials (RCTs) may have unintended consequences on inclusion and cessation disparities. We examined racial/ethnic differences in: a) exclusion from a group-based cessation RCT; and b) reasons for exclusion.Design: Quasi-experimental. Inclusion criteria were self-identification as African American/Black, non-Hispanic White, or Hispanic (any race), adults, minimum five cigarettes/day or carbon monoxide reading of ≥ 8 parts per million (ppm), interest in quitting, and spoke/read English. Data were obtained from a parent trial, which is ongoing and will be completed in 2019. Analyses for our present study on participant screening and enrollment were conducted in 2018.Main Outcome Measures: Study ineligibility, and reasons for exclusion (contraindications for nicotine patch use, serious mental illness [SMI, eg, bipolar disorder or schizophrenia], alcohol dependence or illicit drug use, current tobacco treatment, attendance barriers [eg, transportation], and other concerns [eg, aggressive, intoxicated, disruptive, visibly ill]).Results: Of 1,206 individuals screened, 36% were ineligible. The most frequent reasons were SMI (28%), alcohol dependence or drug use (10%), and attendance barriers (7%). Ineligibility was greater among African Americans (42%) and Hispanics (37%), compared with Whites (24%; P<.001). Compared with African Americans and Hispanics, Whites were more likely to be excluded for single reasons, including attendance barriers, and medical conditions (P<.05). African Americans were more than twice as likely as Whites to be excluded for 3 or more reasons (12% vs 5% respectively, P<.05).Conclusions: A notable proportion of smokers were ineligible for this RCT, with SMI as the greatest single cause. Racial/ ethnic minorities were more likely to be excluded, with African Americans deemed ineligible for multiple reasons. Findings have implications for RCT generalizability, addressing tobacco disparities and health equity.Ethn Dis. 2019;29(1):23-30; doi:10.18865/ed.29.1.23.