Animal models have contributed significantly to our understanding of the underlying biological mechanisms of Alzheimer's disease (AD). As a result, over 300 interventions have been investigated and reported to mitigate pathological phenotypes or improve behavior in AD animal models or both. To date, however, very few of these findings have resulted in target validation in humans or successful translation to disease-modifying therapies. Challenges in translating preclinical studies to clinical trials include the inability of animal models to recapitulate the human disease, variations in breeding and colony maintenance, lack of standards in design, conduct and analysis of animal trials, and publication bias due to under-reporting of negative results in the scientific literature. The quality of animal model research on novel therapeutics can be improved by bringing the rigor of human clinical trials to animal studies. Research communities in several disease areas have developed recommendations for the conduct and reporting of preclinical studies in order to increase their validity, reproducibility, and predictive value. To address these issues in the AD community, the Alzheimer's Drug Discovery Foundation partnered with Charles River Discovery Services (Morrisville, NC, USA) and Cerebricon Ltd. (Kuopio, Finland) to convene an expert advisory panel of academic, industry, and government scientists to make recommendations on best practices for animal studies testing investigational AD therapies. The panel produced recommendations regarding the measurement, analysis, and reporting of relevant AD targets, th choice of animal model, quality control measures for breeding and colony maintenance, and preclinical animal study design. Major considerations to incorporate into preclinical study design include a priori hypotheses, pharmacokinetics-pharmacodynamics studies prior to proof-of-concept testing, biomarker measurements, sample size determination, and power analysis. The panel also recommended distinguishing between pilot 'exploratory' animal studies and more extensive 'therapeutic' studies to guide interpretation. Finally, the panel proposed infrastructure and resource development, such as the establishment of a public data repository in which both positive animal studies and negative ones could be reported. By promoting best practices, these recommendations can improve the methodological quality and predictive value of AD animal studies and make the translation to human clinical trials more efficient and reliable.