One of the major challenges in cancer research is to choose the right type of samples to study. Due to tremendous intratumor heterogeneity within a given primary patient sample, typically a very large cohort of patients is needed in order to reach meaningful conclusions. Other limitations of primary patient samples include their general unsuitability for drug testing, and often limited quantity of cells and/or tissue. Therefore, patient sample-derived cancer cell lines with well-defined genetic and biological characteristics may serve as surrogates for the study of cancer. However, whether these cell lines faithfully represent the true biology of the corresponding cancer is often questioned due to the perception that cell lines may acquire many new mutations during in vitro passaging. By comparing their gene expression profiles, somatic mutation occurrence, and gene copy number, Barretina et al. found that most cancer cell lines correlate very well with their corresponding primary cancers. Furthermore, many of these cell lines do not harbor additional detectable mutations. Most importantly, these cell lines can be used as a modelling tool for testing and predicting new cancer drug sensitivity. This study provides a number of new insights into human cancer. These include the following: 1) plasma cell malignancies, i.e. multiple myeloma, correlate with sensitivity to insulin-like growth factor 1 (IGF-1) receptor inhibitors; 2) aryl hydrocarbon receptor (AHR) expression is associated with MEK inhibitor efficacy in NRAS-mutant lines; and 3) SLFN11 expression predicted sensitivity to topoisomerase inhibitors. In summary, this work identifies a large panel of representative cancer cell lines as irreplaceable tools in our cancer research toolbox.