Metal components subjected to cyclic stress develop surface-evident defects (microcracks, slip bands, etc). Monitoring the formation and evolution of these fatigue damage precursors (FDPs) with increasing numbers of cycles can be an effective tool for determining the fatigue state of the component, which can be used in remaining fatigue life prognostics. In this paper a laser scanning technique for FDP detection is described and experimental results from examination of specimens of several metal types are presented. This technique is based on scanning a focused laser beam over the specimen surface and detecting variations in the characteristics of the scattered light signal. These variations can indicate the presence of surface abnormalities and therefore can be associated with fatigue damage formation. Particular patterns of spatial, angular, and optical characteristics can be used to identify and discriminate many types of FDP, which can provide a means to enhance the accuracy of surface defect frequency estimates and to eliminate the false counts that typically occur on surfaces in uncontrolled environments. Experiments during fatigue testing in the laboratory have shown that the technique can produce a defect frequency estimate that relates well to remaining fatigue life, but previous experiments showed large "plateau" regions, in which the slow defect frequency change made life estimation difficult. New data collection and analysis techniques have therefore been developed, and new experiments have been performed to test the ability of this modified approach to improve the utility of defect frequency measurements over the whole of fatigue life.