Alzheimer's disease (AD) is a neurodegenerative disorder, characterized by a progressive loss of cognitive function. Despite considerable progress, a complete description of the molecular pathology of this disease has yet to be elucidated. In this respect, the need for an animal model that develops some or all aspects of this uniquely human disease in a reproducible fashion is crucial for the development and testing of potential treatments. A valid animal model for AD should exhibit (1) progressive AD-like neuropathology and (2) cognitive deficits, and (3) should be verified in several laboratories. Transgenic models should be able to (4) discern pathogenic effects of familial forms (FAD) mutations from those of transgene overexpression. Models derived from microinjection of FAD mutant alleles should (5) encompass more than one Tg line. At present, however, no model that replicates all of these desirable features exists. In this review, we discuss transgenic mouse models with well-characterized AD-like neuropathology that show some form of cognitive impairment. We argue that conclusions drawn from a limited selection of cross-sectional experiments should be verified in longitudinally designed experiments. Future studies should attempt to establish a closer relationship between molecular pathology and the degree of cognitive impairment. While exact replication of AD in mice may not attainable (due to phylogenetic differences and fundamental differences in behavioral ecology), rigorous comparative analysis of cognitive behavior observed in various mouse models of AD should provide a framework for better understanding of molecular mechanisms underlying cognitive impairment observed in AD patients.