Reaching nonvolunteer female smokers with effective smoking cessation programs is a critical public health challenge. Smokers (N = 2,786) among 15,004 female members of a health maintenance organization who completed a routine needs assessment were invited into the "UCLA Preventive Health Behavior Study," consisting of five telephone interviews over 2 years assessing health practices. Participants (N = 1,396) were randomized into experimental or control conditions of an unsolicited, mailed, self-help smoking cessation program. Subjects were not alerted to the link between the program and the health study. Smoking status was assessed at 1, 6, 12, and 18 months. Across all subjects, point prevalence at 18 months was 18.62, and continuous abstinence was 2.71%. No difference was found between treatment and control groups regarding smoking status or readiness to stop smoking--raising questions about the value of mailing cessation materials to nonvolunteers. Quit rates increased over the 18-month follow-up; those still smoking at 18 months reported increased readiness to quit. Predictors at each follow-up point were examined multivariately.