Since the early genetic studies in yeast, regulation of the cell cycle has been associated with the sequential activation of several proline-directed serine-threonine protein kinases by cyclins. From yeast to humans, the activity of these cyclin-dependent kinases (Cdks) have been thought to be essential for cell cycle regulation. Recent gene-targeted mouse models for different cyclins and Cdks have shown that members of these families show a certain level of redundancy and that specific complexes are not required for the mitotic cell cycle. However, the complexity of the Cdk-cyclin network and the promiscuity of their members makes it difficult to understand the relative contribution of these proteins to the mammalian cell division cycle. Compensatory roles by non-Cdk activities and Cdk-independent functions of cyclins are increasing the complexity of the current simplistic models. We still do not know whether at least one cyclin-dependent kinase activity is required for cell cycle progression in mammalian cells. Indeed, this is a relevant question for cancer therapy.