Predictive models of species distributions are often used in both basic and applied ecology to map species distributions and predict the effects of environmental change. Here, we describe the key concepts relevant to predicting species distributions (focusing on the use of niche theory), the types of data typically used, some common modeling algorithms, and illustrate how models are frequently evaluated. Our general goal is to illustrate how concepts, data, and models are used to interpret species–environment relationships and create maps of species distributions for addressing ecological questions and conservation problems. To do so, we model the distribution of the varied thrush (Ixoreus naevius) in the western United States using several model algorithms, such as climate envelopes, generalized linear and additive models, random forests, and Maxent. This example illustrates the different assumptions of modeling algorithms and how understanding the utility of models can vary based on how models are evaluated. Finally, we link these diverse approaches by emphasizing how many of these approaches can be cast as approximations of inhomogeneous point process models, which can help guide modeling decisions. We end by discussing further advances in applied modeling of species distributions that aim to improve predictions, account for measurement error, and incorporate dynamics into the modeling process.