The Golgi apparatus is composed of biochemically distinct early (cis, medial) and late (trans, TGN) cisternae. There is debate about the nature of these cisternae. The stable compartments model predicts that each cisterna is a long-lived structure that retains a characteristic set of Golgi-resident proteins. In this view, secretory cargo proteins are transported by vesicles from one cisterna to the next. The cisternal maturation model predicts that each cisterna is a transient structure that matures from early to late by acquiring and then losing specific Golgi-resident proteins. In this view, secretory cargo proteins traverse the Golgi by remaining within the maturing cisternae. Various observations have been interpreted as supporting one or the other mechanism. Here we provide a direct test of the two models using three-dimensional time-lapse fluorescence microscopy of the yeast Saccharomyces cerevisiae. This approach reveals that individual cisternae mature, and do so at a consistent rate. In parallel, we used pulse-chase analysis to measure the transport of two secretory cargo proteins. The rate of cisternal maturation matches the rate of protein transport through the secretory pathway, suggesting that cisternal maturation can account for the kinetics of secretory traffic.