Background: Tobacco dependence is a chronic condition requiring repeat treatment and long-term follow-up. However, current evidence regarding best approaches to providing these services is fairly limited. This study assessed the effectiveness of an intervention for increasing repeat tobacco-dependence treatment in a population of relapsed smokers. Design: The study was a multicenter randomized controlled trial. Setting/participants: Patients aged 19 and older receiving a prescription for nicotine replacement therapy or bupropion for smoking cessation in 2002 at one of five participating Veterans Affairs (VA) facilities were eligible. Intervention: A total of 1900 eligible veterans were randomized to (1) patient phone call to assess smoking status, quit challenges, and treatment preferences, and computerized progress note to providers communicating this information (intervention); or (2) usual care (control). Main outcome measures: Outcomes, assessed in 2003, included the proportion of patients receiving repeat pharmacologic or behavioral smoking-cessation treatment in the 6-month follow-up period (assessed from VA pharmacy and outpatient data), 7-day point prevalence abstinence, and satisfaction (assessed by patient survey). Results: Thirty-four percent of intervention participants versus 22% of controls received treatment for tobacco dependence in the follow-up period (p =0.0001). Thirty-two percent of intervention participants versus 21% of controls received pharmacologic treatment (p =0.0001), and 12% of intervention subjects versus 8% of controls received behavioral treatment (p =0.005). Abstinence rates at follow-up were 25% for intervention subjects and 22% for controls (p =0.15). Intervention subjects were more likely than controls to report being satisfied with the general smoking-cessation help (87% vs 82%, p =0.01), as well as the pharmacologic help that they received from the VA (89 vs 84%, p =0.01). Conclusions: The intervention significantly increased repeat treatment rates and satisfaction with services but did not have a significant effect on abstinence rates.