As cancer immunotherapies become mainstream for the treatment of many different cancer types and the numbers of new agents continue to increase, the need for experimental models is also rising. An approach to develop and study models for immune-oncology that has garnered intense interest in recent years is that of using patient-derived models. Patient-derived models can recapitulate many of the features and heterogeneity of the corresponding human tumors. Historically these models have been used to study cancer cell-intrinsic properties of tumors and drugs that target tumor cells directly. In recent years, increasing recognition of the role immune cells play in cancer and how these represent good therapeutic targets has led to efforts to optimize and use patient-derived models for cancer immunotherapy studies. Patient-derived models are now being used to study the properties of cancer cells that modulate their ability to respond to immune stimulation. Further efforts are underway to use and develop patient-derived models that incorporate human immune cells in vitro and in vivo (humanized mice) so that cancer cell-immune cell interactions can be studied in the context of cancer immunotherapies. As these models are further refined, leveraging patient-derived models for cancer immunotherapy research can provide insight into mechanisms of sensitivity and resistance to cancer immunotherapies, uncover new targets, reveal how specific agents work, and be used to evaluate the antitumor efficacy of therapeutic regimens.