The debate on the cell of origin of human retinoblastoma lasted for more than one century. In the recent issue of ldquo;cellrdquo;, David Cobrinikprime;s group shows that L/M cone precursors are the most likely answer as they have an intrinsic circuitry, including murine double minute 2 (MDM2), Nmyc, the nuclear receptors retinoid X receptor and thyroxine receptor 2, making them extremely sensitive to transformation following retinoblastoma gene inactivation.