The purpose of this paper is threefold: (1) to identify gaps in the literature and inconsistencies between empirical data and Rogers' diffusion of innovations model originally outlined in 1962; (2) to propose a model of fashion adoption; and (3) to conduct an empirical test of the proposed model. The proposed model varies from Rogers' model as follows: four (vs. five) ideal‐type adopter groups with unique names and clear operational definitions; proportional distribution of consumers among groups; clear method of statistically identifying group members; categorization based on fashion innovativeness and opinion leadership; and standard procedure for identifying groups based on mean and standard deviation. The proposed model offers greater precision and provides results that are comparable across studies. The model was analyzed empirically using four methods of categorizing participants into fashion adoption groups (independent variable) and fashion involvement as the dependent variable. Participants in the study were 309 students from two universities. The analysis using the proposed model provided clear‐cut results that supported the hypothesis that higher levels of consumers' fashion involvement increased the probability of earlier adoption. Results from the other three analyses were not as clear‐cut. Based on the model for fashion adoption presented in this paper, the foundation is primed for development of a theory of fashion adoption. The proposed model offers a solution to the problem of identifying fashion adoption groups, plus, the solution achieves the desired effect with the simplest effort and is undertaken with appropriate methods.