ABSTRACTWhile a fairly large body of literature dealing with just‐in‐time (JIT) manufacturing exists, most of it is descriptive. JIT research consists primarily of case studies, surveys, a few analytic models (primarily dealing with Kanban), and some simulations. This research builds on the existing literature to develop and test an exploratory model of JIT purchasing with an empirical field study of an actual JIT implementation. While the results basically follow hypotheses developed from the literature, some anomalies were found. The research provides both a methodology for further JIT study and an indication of the possible relative importance of JIT purchasing characteristics as they exist in a North American installation. Especially insightful is the fact that the relative importance changes depending on whether customer inventory or supplier inventory is used as the dependent variable. With customer inventory, the variables measuring delivered lot size, the size of release changes, and supplier manufacturing lead time were significant. With supplier inventory, only supplier manufacturing lot size was signific