Significance Many new cancer drugs fail at the clinical stage owing to poor efficacy and/or excessive toxicity, though whether this reflects shortcomings of the target or the drug is often unclear. To gain earlier insights into factors that can influence the therapeutic index of target inhibition in vivo, we combine inducible RNA interference and somatic engineering technologies to produce a cost-effective platform that enables systemic and inducible suppression of candidate target in normal tissues and tumor cells in the same mouse. By comparing the consequences of genetic and pharmacological CDK9 inhibition, we establish the utility of this platform to predict factors influencing the therapeutic index. Additionally, our studies provide support, and some cautionary notes, for the clinical development of CDK9 inhibitors.