OBJECTIVE Adenosine modulates neurotransmission and in the intestine adenosine is continuously released both from nerves and from smooth muscle. The main effect is modulation of contractile activity by inhibition of neurotransmitter release and by direct smooth muscle relaxation. Estimation of adenosine concentration at the receptors is difficult due to metabolic inactivation. We hypothesized that endogenous adenosine concentrations can be calculated by using adenosine receptor antagonist and agonist and dose ratio (DR) equations. METHODS Plexus-containing guinea-pig ileum longitudinal smooth muscle preparations were made to contract intermittently by electrical field stimulation in organ baths. Schild plot regressions were constructed with 2-chloroadenosine (agonist) and 8-(p-sulfophenyl)theophylline (8-PST; antagonist). In separate experiments the reversing or enhancing effect of 8-PST and the inhibiting effect of 2-chloroadenosine (CADO) were analysed in the absence or presence of an adenosine uptake inhibitor (dilazep), and nucleoside overflow was measured by HPLC. RESULTS Using the obtained DR, baseline adenosine concentration was calculated to 28 nm expressed as CADO activity, which increased dose dependently after addition of 10(-6) m dilazep to 150 nm (P < 0.05). HPLC measurements yielded a lower fractional increment (80%) in adenosine during dilazep, than found in the pharmacological determination (440%). CONCLUSIONS Endogenous adenosine is an important modulator of intestinal neuro-effector activity, operating in the linear part of the dose-response curve. Other adenosine-like agonists might contribute to neuromodulation and the derived formulas can be used to calculate endogenous agonist activity, which is markedly affected by nucleoside uptake inhibition. The method described should be suitable for other endogenous signalling molecules in many biological systems.