OBJECTIVE: This study aimed to compare the effectiveness of two cognitive behavioral therapy-based smoking cessation interventions initiated during hospitalization and to evaluate the factors related to relapse after discharge. METHODS: This was a prospective randomized study involving 90 smokers hospitalized in a university hospital. We collected data related to sociodemographic characteristics, reasons for admission, smoking-related diseases, smoking history, the degree of nicotine dependence (ND), and the level of craving. Patients were divided into two treatment groups: brief intervention (BrInter, n = 45); and intensive intervention with presentation of an educational video (InInterV, n=45). To assess relapse, all patients were assessed by telephone interview in the first, third, and sixth months after discharge. Abstinence was confirmed by measurement of exhaled carbon monoxide (eCO). RESULTS: Of the 90 patients evaluated, 55 (61.1%) were male. The mean age was 51.1 ± 12.2 years. The degree of ND was elevated in 39 (43.4%), and withdrawal symptoms were present in 53 (58.9%). The mean eCO at baseline was 4.8 ± 4.5 ppm. The eCO correlated positively with the degree of ND (r = 0.244; p = 0.02) and negatively with the number of smoke-free days (r = -0.284; p = 0.006). There were no differences between the groups in terms of the variables related to socioeconomic status, smoking history, or hospitalization. Of the 81 patients evaluated at 6 months, 33 (40.7%) remained abstinent (9 and 24 BrInter and InInterV group patients, respectively; p = 0.001), and 48 (59.3%) had relapsed (31 and 17 BrInter and InInterV group patients, respectively; p= 0.001). Moderate or intense craving was a significant independent risk factor for relapse, with a relative risk of 4.0 (95% CI: 1.5-10.7; p < 0.00001). CONCLUSIONS: The inclusion of an educational video proved effective in reducing relapse rates. Craving is a significant risk factor for relapse.