In an extension and replication of previous work on social support in worksite smoking programs, 29 employees were assigned to either a basic smoking control program or to a basic treatment plus significant other support condition. Within a multiple baseline across behaviors design, all subjects received a 6 week treatment program that focused on achieving sequential reductions in nicotine content of brand smoked, number of cigarettes smoked per day, and percent of the cigarette smoked. Both treatment conditions were equally successful in producing abstinence (verified by biochemical analyses) and in producing reductions in smoking behavior among nonabstinent subjects at both posttest and 6-month follow-up assessments. In contrast to previous research with this program, there was considerable relapse in both conditions by follow-up. Consistent with previous findings, supportive social interactions were not related to treatment outcome, but the level of negative (nonsupportive) social interactions was inversely correlated with treatment success. Implications of these findings and directions for future research are discussed.