Alzheimer's disease (AD) is the most common form of senile dementia, and a cure is desperately needed. The amyloid-beta42 (Abeta42) has been suggested to play a central role in the pathogenesis of AD. However, the mechanism by which Abeta42 causes AD remains unclear. To understand the pathogenesis and to develop therapeutic avenues, it is crucial to generate animal models of AD in genetically tractable organisms. Drosophila is a well-established model system for which abundant genetic tools are available. Moreover, its well organized brain permits the study of complex behaviors such as learning and memory. We have established transgenic flies that express human Abeta42 in the nervous system. These flies developed age-dependent short-term memory impairment and neurodegeneration. In this review, we will first describe transgenic Abeta42 fly models and discuss the unique features of this system compared to mouse AD models. Secondly, we will discuss the usage of the fly models to evaluate currently proposed therapeutic strategies. Thirdly, we will briefly review the results of a genetic screen for modifiers of Abeta42 toxicity in the fly model. Finally, we will discuss how to dissect the complex mechanisms of Abeta42 toxicity focusing on its aggregation propensity using the fly model system.