Just-in-time (JIT) manufacturing systems have found wide acceptance in the automotive industry as well as in other industries such as electronic assembly. One of the key features of JIT systems is frequent reliable deliveries from local vendors. This article considers two major points related to managing inbound transportation from these local vendors. The first point is that vendor delivery uncertainties can be reduced if the JIT manufacturer takes over the responsibility for inbound logistics from all JIT vendors. In other words, we are suggesting that the manufacturing firm (vendee) pick up raw materials and purchased parts from its vendors rather than having the vendors deliver. The second major point is that a simple economic analysis and a computer-based decision support system (like that we propose in this article) could be used to help schedule the JIT pickups. The article begins with a description of why deliveries are critical in the JIT environment and why vendee pickup from local vendors can be a better alternative than vendor delivery. When a manufacturer manages its own inbound local deliveries it benefits from more timely information on upsets, reduced transportation costs, reduced transaction costs, and most importantly, reduced uncertainty in deliveries. We then turn to a model for how to schedule JIT pickups. This model addresses the following "JIT Vendor Pickup Problem": For a given number of vehicles with limited capacity, find a one-week JIT vendor pickup schedule that will minimize the total incremental carrying, travel, and pickup costs subject to the constraint that the vendee does not run out of any item from any vendor. In the above problem definition, the demand rates for all items are assumed to be constant over the short term. This model can be imbedded in a computer-based decision support system to facilitate the daily scheduling of the pickups from the vendors. In the conclusion, we deal with the expected benefits to the vendee for managing its own incoming JIT pickups and then propose a model that could be used to implement this concept.