A portacaval shunt (PCS) model is frequently employed to study phenomena inherent in portal-systemic shunting of splanchnic blood. In many species, a PCS induces hepatic insufficiency, accompanied by encephalopathy. Rats operated on with a 'nonsuture' technique tolerate a PCS better and exhibit no or only slight encephalopathy. Age and environment seem to have a large impact on the ability to tolerate a PCS. This explains the discrepancies between the results of different investigators and the varying time periods reported between the PCS operation and the optimal time for experiments. To characterize the PCS model (button technique) in rats with respect to metabolic parameters in our field of interest, we studied three groups of male Sprague-Dawley rats--non-operated (n = 12); sham-operated (n = 12) and PCS (n = 13)--for 4 weeks following surgery. Body weight in the PCS group decreased for 1 week after surgery and then increased at about the same rate as in the control groups. Plasma immunoreactive insulin, plasma immunoreactive glucagon (IRG) and aromatic amino acid concentrations were highest 1 week after surgery and tended to normalize in the next weeks. Plasma branched-chain amino acid (BCAA) concentrations were decreased in the 1st, 2nd and 3rd week after surgery, after which normalization occurred. These data demonstrate that after 3-4 weeks, male Sprague-Dawley rats start to recover from the metabolic disturbances caused by PCS with regard to the parameters measured. Therefore, experiments in this area, especially those relating to BCAA metabolism, should be carried out 2-3 weeks after the shunt operation (button technique).