Several authors have shown that angiotensin II stimulates hepatic angiotensinogen synthesis in vivo, ex vivo and in vitro. In previous studies we have demonstrated that this effect of angiotensin II depends mainly on a transient inhibition of adenylyl cyclase and is the consequence of a stabilization of angiotensinogen mRNA. In the present study we describe the isolation of a polysomal 12 kDa protein which, in band shift and cross link assays, shows a specific affinity to the 3′ untranslated region (3′ UTR) of angiotensinogen mRNA and prevents enzymatic degradation of angiotensinogen mRNA in a cell-free incubation system. [32P]UTP-labelled or unlabelled 3′ fragments of angiotensinogen mRNA were synthesized on a transcription vector (pGEM5zf+) into which the corresponding DNA sequence was cloned after restriction from vector pRAG 16. Binding of the 12 kDa protein to the radioactively labelled 3′ UTR of angiotensinogen mRNA could be displaced by unlabelled 3′ UTR mRNA fragments but not by a renin mRNA of comparable length derived from the coding region, The RNA-binding protein appears to be derived from a higher molecular mass precursor (45 kDa) which is preferentially present under reducing conditions in vitro; the active low molecular mass form is evident in the absence of reducing agents. In a cross link experiment we established that a band shift signal which was obtained in the presence of the 45 kDa protein preparation exclusively depends on RNA binding of the active 12 kDa protein. In addition, a phosphorylation step may be involved in the activation of the 12 kDa protein, since its molecular mass and isoelectric point correlate with proteins which were phosphorylated in response to transient decreases of cAMP (induced by guanfacine or angiotensin II) or in response to a direct inhibition of protein kinase A by the cAMP antagonist Rp-cAMP. The importance of phosphorylation reactions for the stabilization of angiotensinogen mRNA was further assessed in a cell-free incubation system of rat liver parenchymal cells. These studies demonstrated that in the presence of acid phosphatase (1 U/ml) the half-life of angiotensinogen was significantly decreased. In the same incubation system the 12 kDa protein increased the half-life of endogenous as well as of exogenous angiotensinogen mRNA three- to fourfold, while no stabilizing effect was apparent when exogenous angiotensinogen mRNA from which the 3′ tail had been deleted was added. We concluded that an intracellular 12 kDa protein may play a crucial role in the angiotensin Il-induced stabilization of hepatic angiotensinogen mRNA and further suggest that this protein exerts its effect via binding to the 3′ UTR of angiotensinogen mRNA in response to a cAMP-dependent activation step.