Forward Collision Alert (or Forward Collision Warning) systems provide alerts intended to assist drivers in avoiding or mitigating the harm caused by rear-end crashes. These systems currently use front-grille mounted, forward-looking radar devices as the primary sensor. In contrast, Lane Departure Warning (LDW) systems employ forward-looking cameras mounted behind the windshield to monitor lane markings ahead and warn drivers of unintended lane violations. The increasing imaging sensor resolution and processing capability of forward-looking cameras, as well recent important advances in machine vision algorithms, have pushed the state-of-the-art for camera-based features. Consequently, camera-based systems are emerging as a key crash avoidance system component in both a primary and supporting sensing role. There are currently no production vehicles with cameras used as the sole FCA sensing device. This paper discusses the development of a camera-based FCA system that uses a camera in place of a radar device for sensing rear-end crash situations. This paper provides an overview of the system, including how the system detects vehicles, tracks vehicles, projects collision course trajectories, and estimates time-to-collision (TTC) using image scale change. Results from test track and public road testing support the deployment of a camera-based FCA system, and indicate this system would meet the United States Department of Transportation New Car Assessment Program (NCAP) Forward Collision Warning confirmation test requirements. Furthermore, the conditions under which most rear-end crashes occur suggests that this system provides a promising approach to reduce the harm caused by rear-end crashes.