A stage-matched expert system intervention was evaluated on 4144 smokers in a two-arm randomized control trial with four follow-ups over 24 months. Smokers were recruited by random digit-dial calls, and 80.0% of the eligible smokers were enrolled. Individualized and interactive expert system computer reports were sent at 0, 3, and 6 months. The reports provided feedback on 15 variables relevant for progressing through the stages. The primary outcomes were point prevalence and prolonged abstinence rates. At 24 months, the expert system resulted in 25.6% point prevalence and 12% prolonged abstinence, which were 30% and 56% greater than the control condition. Abstinence rates at each 6-month follow-up were significantly greater in the Expert System (ES) condition than in the comparison condition with the absolute difference increasing at each follow-up. A proactive home-based stage-matched expert system smoking cessation program can produce both high participation rates and relatively high abstinence rates.