Abstract c-Met is a receptor tyrosine kinase with oncogenic potential for which several small molecule inhibitors are currently being tested in clinical trials. There are suggestions in the literature that c-Met inhibitors could have anti-tumor effects under a variety of contexts, including in tumors in which the MET gene is amplified, contains a gain-of-function mutation or in tumors with receptor overexpression. We have developed ATP-competitive c-Met inhibitors that possess different selectivity profiles and fall into two categories. Class I molecules are highly selective for c-Met while class II molecules are multi-kinase inhibitors. These molecules were profiled in vitro and in vivo to identify biomarkers of c-Met dependence. We show that both classes of molecules inhibited c-Met kinase activity in vitro and in vivo. However, class I and class II molecules differed extensively in their spectrum of anti-tumor activity. The anti-proliferative effects of the different c-Met inhibitors were tested in 359 cancer cell lines in vitro. The activity of class I molecules was exclusively restricted to cell lines that harbor MET amplification. Western blotting in sensitive or resistant cell lines showed that c-Met inhibitors effectively block signaling downstream of c-Met only in MET amplified cell lines. In vivo, class I molecules inhibited only the growth of tumor models that are highly dependent on c-Met activity due to MET amplification or due to a hepatocyte growth factor (HGF)-driven autocrine loop. The same models that were sensitive to class I inhibitors were also sensitive to class II inhibitors. However, class II inhibitors had additional anti-tumor activities in vitro and in vivo in models that did not respond to class I inhibitors. The majority of cell lines that responded only to a class II molecule in vitro required much higher drug concentrations than those required to inhibit the growth of MET amplified cell lines. Finally, the growth inhibitory effects observed in vitro and in vivo were consistent with the selectivity profiles of the different c-Met inhibitors. Together our data demonstrate that MET amplification is a clinically identifiable, potential sensitivity biomarker for selective c-Met kinase inhibitors. c-Met expression alone is not a sensitivity biomarker in preclinical models. The spectrum of c-Met inhibitor sensitive models defined in this work could be used to understand whether other small molecule c-Met inhibitors are exquisitely selective for c-Met or harbor additional activities. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 3558. doi:10.1158/1538-7445.AM2011-3558