This paper addresses the active and dynamic nature of biometrics, in general, and gaze analysis, in particular, including motivation and background. The paper includes a critical survey of existing gaze analysis methods, challenges due to uncontrolled settings and lack of standards, and outlines promising future R&D directions. Criteria for performance evaluation are proposed, and state-of-the art gaze analysis methods are compared on the same database set. Performance improvement would come from richer stimuli including task dependent user profiles, with applications going much beyond identity management to include personalized medical care and rehabilitation, privacy, marketing, and education.