Recent theoretical studies of flaked stone technology have identified many factors that affect the ways in which human beings make and use tools. However, these studies lack a unified body of theory that might help to integrate their diverse perspectives. This paper expands recent anthropological discussions of risk as the basis for such a theory. We begin by defining risk, emphasizing two distinct components of this concept-the probability that some problem will occur and the cost of such an occurrenceand argue that technology can be seen as a means of reducing such probabilities in the face of unacceptably high costs. We support this argument using cross-cultural data on hunter-gatherer technology and archaeological and historic data on the construction of defensive works on the northern Great Plains. Next, we consider specific problems in applying this perspective archaeologically, concluding that existing limits on our ability to estimate failure probabilities and costs prevent us from testing the ideas outlined here in archaeological contexts. However, accepting this perspective as provisionally validated by ethnographic data allows us to see how it can illuminate archaeological cases, and we exemplify this by comparing the production of Araya microblade projectile points in the Japanese Paleolithic and Folsom fluted projectile points on the North American Great Plains.