BACKGROUND: The nicotine patch has been widely used for smoking cessation, but not all smokers quit using the patch. Knowing which smokers are likely to succeed with the nicotine patch may improve the efficiency of nicotine patch use. OBJECTIVE: To identify predictors of smoking abstinence using baseline characteristics, smoking behavior, and withdrawal symptoms. METHODS: Using 2 randomized clinical trials of pharmacologic treatment, brief counseling, and quit date formats in the outpatient research clinic setting, predictors of smoking cessation were derived in 1 sample (n = 159), then prospectively validated in an independent sample (n = 48). Subjects smoked 1 pack of cigarettes per day or more and were motivated to quit smoking. Self-report of abstinence at 6 months verified by exhaled carbon monoxide of 8 ppm or less was used. RESULTS: Abstinence at 6 months was 24% in the derivation set and 25% in the validation set. Using logistic regression, a model containing quit date abstinence (odds ratio, 10.6; 95% confidence interval [CI], 2.9-38.7) and baseline nicotine dependence (odds ratio, 0.75; 95% CI, 0.6-1.0 per unit increase in Fagerstrom score) provided the optimal predictive ability and was validated in the independent set. Quit date abstinence improved the likelihood of 6-month abstinence by 4.1 over baseline (95% CI, 2.6-6.4) for low-nicotine-dependent smokers and 1.2 (95% CI, 0.6-2.2) for high-nicotine-dependent smokers. Quit date smoking altered the likelihood of 6-month abstinence by 0.2 (95% CI, 0.0-0.8) for low-dependent smokers and 0.1 for high-dependent smokers (95% CI, 0.0-0.6). CONCLUSIONS: Abstinence on the quit date and low-nicotine dependence improve the likelihood of smoking abstinence at 6 months. Smoking on the quit date may be an indication for postponing the cessation attempt or adjusting the therapy for smoking cessation.