Systematic biases in cross-sectional community studies may underestimate the effectiveness of stop-smoking medications.