Falk and Dierking's Contextual Model of Learning was used as a theoretical construct for investigating learning within a free-choice setting. A review of previous research identified key variables fundamental to free-choice science learning. The study sought to answer two questions: (1) How do specific independent variables individually contribute to learning outcomes when not studied in isolation? and (2) Does the Contextual Model of Learning provide a useful framework for understanding learning from museums? A repeated measure design including interviews and observational and behavioral measures was used with a random sample of 217 adult visitors to a life science exhibition at a major science center. The data supported the contention that variables such as prior knowledge, interest, motivation, choice and control, within and between group social interaction, orientation, advance organizers, architecture, and exhibition design affect visitor learning. All of these factors were shown to individually influence learning outcomes, but no single factor was capable of adequately explaining visitor learning outcomes across all visitors. The framework provided by the Contextual Model of Learning proved useful for understanding how complex combinations of factors influenced visitor learning. These effects were clearerest when visitors were segmented by entry conditions such as prior knowledge and interest.