The tumour suppressor PTEN, also named MMAC1 or TEP1, is associated with a number of malignancies in human populations. This protein has a dual protein phosphatase activity, being also capable to dephosphorylate phosphatidylinositol 3,4,5 triphosphate. We have studied the mechanism of growth suppression attributable to PTEN. We observed that PTEN overexpression inhibits cell growth in a variety of normal and transformed, human and murine cells. Bromodeoxyuridine (BrdU) incorporation and TUNEL labelling experiments in transiently transfected cells demonstrate that this inhibition is due to a cell cycle arrest rather than induction of apoptosis. Given that PTEN is unable to cause cell growth arrest in retinoblastoma (Rb)-deficient cell lines, we have explored the possible requirement for pRb in the PTEN-induced inhibition of cell proliferation. We found that the co-expression of SV40 antigen, but not a mutant form (which binds exclusively to p53), and cyclin D1/cdk4 are able to overcome the PTEN-mediated growth suppression. In addition, the reintroduction of a functional pRb, but not its relatives p107 or p130, in Rb-deficient cells restores the sensitivity to PTEN-induced arrest. Finally, the hyperphosphorylation of transfected pRb is inhibited by PTEN co-expression and restored by PI-3K co-expression. Accordingly, PTEN gene is mostly expressed, in parallel to Akt, in mid-late G1 phase during cell cycle progression prior to pRb hyperphosphorylation. Finally, we have studied the signal transduction pathways modulated by PTEN expression. We found that PTEN-induced growth arrest can be rescued by the co-expression of active PI-3K and downstream effectors such as Akt or PDK1, and also certain small GTPases such as Rac1 and Cdc42, but not by active Ha-ras, raf or RhoA. Collectively, our data link the tumour suppressor activities of PTEN to the machinery controlling cell cycle through the modulation of signalling molecules whose final target is the functional inactivation of the retinoblastoma gene product.