This study evaluates four measures of exposure to retail cigarette marketing in relation to adolescent smoking behavior. The measures are (a) shopping frequency in types of stores known to carry more cigarette advertising than other store types, (b) shopping frequency in specific stores that sell cigarettes in the study community, (c) the amount of exposure to cigarette brand impressions in stores where students shopped, and (d) perceived exposure to cigarette advertising. The study combined data from classroom surveys administered to 6th-, 7th-, and 8th-grade students in three California middle schools, and direct store observations quantifying cigarette marketing materials and product placement in stores where students shopped. Logistic regression models were used to examine how each exposure measure related to the odds of ever smoking and susceptibility to smoke, controlling for grade, gender, ethnicity, school performance, unsupervised time, and exposure to household and friend smoking. Frequent exposure to retail cigarette marketing as defined by each of the four measures was independently associated with a significant increase in the odds of ever smoking. All but the measure of exposure to store types was associated with a significant increase in the odds of susceptibility to smoke. Four measures of exposure to retail cigarette marketing may serve equally well to predict adolescent smoking but may vary in cost, complexity, and meaning. Depending on the outcomes of interest, the most useful measure may be a combination of self-reported exposure to types of stores that contain cigarette marketing and perceived exposure to such messages.