How can the manufacturability of different product design alternatives be evaluated efficiently during the early stages of concept exploration? The benefits of such integrated product and manufacturing process design are widely recognized and include faster time to market, reduced development costs and production costs, and increased product quality. To reap these benefits fully, however, one must examine product/process trade-offs and cost/schedule/performance trade-offs in the early stages of design. Evaluating production cost and lead time requires detailed simulation or other analysis packages which 1) would be computationally expensive to run for every alternative, and 2) require detailed information that may or may not be available in these early design stages. Our approach is to generate response surfaces that serve as approximations to the analyses packages and use these approximations to identify robust regions of the design space for further exploration. In this paper we present a method for robust product and process exploration and illustrate this method using a simplified example of a machining center processing a single component. We close by discussing the implications of this work for manufacturing outsourcing, designing robust supplier chains, and ultimately designing the manufacturing enterprise itself.