Activation of cell surface growth factor receptor tyrosine kinases (RTKs) results in cellular proliferation, differentiation and survival. Humans have 58 RTKs, categorized into twenty subfamilies based on their structural properties and ligand specificity. One of the most well-studied RTKs, the epidermal growth factor receptor (EGFR), is mutated or overexpressed in many human cancers including non-small cell lung cancer and glioblastoma. Several EGFR inhibitors have been approved by the FDA, and numerous additional inhibitors are being pursued as potential cancer therapeutics. Structural studies of the EGFR tyrosine kinase domain suggest that different inhibitors bind selectively to either the active or the inactive conformation of the kinase. Contrary to this supposition, computational studies of inhibitor-stabilized EGFR conformations have suggested that erlotinib can bind equally well to either the active or inactive forms of the EGFR kinase domain. We tested this hypothesis using a mutated EGFR kinase domain that can adopt either the inactive or active kinase conformation in crystals, and determined a structure with erlotinib bound to the inactive conformation – suggesting that this inhibitor can bind either ‘state’. We also determined the crystal structure of this EGFR kinase variant bound to a novel ErbB2 inhibitor (55A) found in screens designed to select inhibitors that bind the active state. When bound to this compound, the EGFR kinase domain adopted only the active conformation, suggesting that erlotinib and 55A select different kinase ‘states’. Collaborative approaches combining biochemical, structural and cell signaling studies suggest new considerations in predicting modes of inhibitor binding in the development of therapeutic agents for cancer. Several activating mutations in the anaplastic lymphoma kinase (ALK) gene have been implicated in neuroblastoma. The small molecule tyrosine kinase inhibitor crizotinib, which inhibits ALK and Met, was recently approved by the FDA for the treatment of non-small cell lung cancer (NSCLC), and is currently in early-phase clinical testing in patients with neuroblastoma (NB). Numerous additional ALK inhibitors are also being pursued as potential therapeutics. However, more and more mutations in the ALK tyrosine kinase domain (TKD) are being found in patients with neuroblastoma. Several of these are associated with primary resistance to currently available ALK inhibitors, and several mutations in oncogenic ALK fusions have been linked to acquired crizotinib resistance. Understanding how different ALK mutations activate its kinase domain, and how they change inhibitor sensitivity, is therefore crucial for further clinical development of ALK-targeted therapeutics. We have studied the biochemical consequences of a wide variety of ALK kinase domain mutations in parallel both with studies of their transforming abilities and computational studies of their structural effects. Our first goal was to use these data to predict the effects of new clinically emerging ALK mutations on ALK’s transforming ability, signaling activity, and drug sensitivity. A second goal was to identify the most potent ALK inhibitor that has inhibitory effects on the widest possible range of ALK mutants using biochemical and cellular assays, with a view to identifying a clinical path forward. Our studies provide valuable mechanistic insight into how ALK is regulated, and also lay important groundwork for guiding more refined targeted therapy in neuroblastoma patients.