Cancer is a remarkably diverse illness, exhibiting heterogeneity both within a patient (intratumoral heterogeneity) and between patients (intertumoral heterogeneity). The lack of preclinical models, which accurately represent the variety of tumors in people, is one of the most stated causes of the high failure rate of novel oncology treatments. The development of patient-derived tumor xenograft (PDX) is a potent technology that can preserve, in major part, molecular heterogeneity of the patient tumor. The genetic and histological features of human patient tumors are easily maintained in PDX models since they are created from patient tumor tissues. As a result, PDX models can be used at several phases of the development of cancer diagnostics and treatments, such as co-clinical trials, drug efficacy testing, biomarker development, companion diagnostics, and overcoming drug resistance. The many facets of PDX models are covered in this chapter, along with their use in cancer research, successes of PDX models for different types of cancer, difficulties facing the current generation of PDXs, and a summary of the models’ future directions.