Understanding the relationship between triplicated genes and Down syndrome (DS) phenotypes, including cellular and molecular mechanisms leading to these traits, has been facilitated through the use of DS mouse models. Mouse models provide an indispensable resource to understand geneotype–phenotype relationships in Trisomy 21 (Ts21). Experiments utilizing DS mouse models have produced key results addressing the molecular pathophysiology of DS. DS mouse models are useful in better understanding Ts21 and its resultant attributes, as well as predicting potential treatments. Although they do not exactly recapitulate human genetics or phenotypes, mouse models share functional mechanisms with humans. An overall best DS mouse model for Ts21 and every DS-related phenotype does not exist; instead, an array of different models are available with unique features to test hypotheses and understand more about DS. DS mouse models include different genetic constructs to create dosage imbalance to model Ts21. Phenotypic variability is apparent in individuals with Ts21 and is also observed in DS mouse models; this variation must be accounted for when utilizing DS mouse models. Because of differing genetic composition and Ts21-related phenotypes in DS mouse models, the ability to represent traits accurately and to predict how results from animal models may represent the human condition may vary between models. Therefore, it is important to understand DS mouse models, their individual benefits, and limitations, and select those models that have the highest probability for success in representing a particular DS-related condition. Potential confounding factors, including presence or absence of an extra chromosome, differences between sexes, genetic background, and phenotypic modifiers should be considered as DS mouse models are selected to test the chosen hypothesis. The following discussion of these issues will be useful in selecting DS mouse models, maximizing the knowledge generated using these mice, and better understanding the genotype–phenotype relationship in DS.