TGF-beta is a potent inhibitor of cell growth, and accumulating evidence suggests that perturbation of the TGF-beta signaling pathway leads to tumorigenesis. Smads are recently identified proteins that mediate intracellular signaling of the TGF-beta superfamily. Smads 2 and 3 are phosphorylated by the TGF-beta type I receptor. Smad4 was originally identified as a candidate tumor suppressor gene in pancreatic cancers. Smads 2 and 3 form complexes with Smad4 upon TGF-beta stimulation. The heteromeric Smad complexes translocate into the nucleus, where they activate expression of target genes. Our recent study demonstrated that Smads exist as monomers in the absence of TGF-beta. Smads 2 and 3 form homo- as well as hetero-oligomers with Smad4 upon ligand stimulation. Both homo-oligomers and hetero-oligomers directly bind to DNA, suggesting that the signaling pathway of Smads may be multiplex. Smads 2 and 3 associate with transcriptional coactivators such as p300 in a ligand-dependent manner, p300 enhances transactivation by TGF-beta, suggesting that coactivators link Smads to the basal transcriptional machinery. A missense mutation of Smad2 identified in colorectal and lung cancers was introduced to Smad3. The mutant, Smad3(DE), blocked the activation of wild-type Smad2 and Smad3. Thus, the missense mutation not only disrupts the function of the wild-type Smad but also creates a dominant-negative Smad, which could actively contribute to oncogenesis.