1. In the absence of adenosine uptake inhibition, adenosine produced a concentration-dependent (threshold 30 microM) relaxation of the 5-methylfurmethide pre-contracted guinea-pig taenia caecum. The relaxation was not blocked by 8-phenyltheophylline (8-PT, 3 microM) or 1,3-dipropyl, 8-cyclopentylxanthine (DPCPX, 30 microM). 2. In the presence of the adenosine uptake inhibitor, dipyridamole (Dip, 3 microM), a biphasic adenosine concentration-effect curve was obtained (threshold 0.3 microM). The time course of the responses to adenosine in the absence of Dip was similar to that of the second phase responses in the presence of Dip and occurred over the same adenosine concentration-range. 5'-(N-ethyl) carboxamido-adenosine (NECA) concentration-effect curves (in the absence of Dip) were also biphasic. Only the first phases of the concentration-effect curves obtained with NECA and adenosine (plus Dip) were inhibited by 8-PT. The pA2 values for 8-PT of 6.7 and 7.0 versus adenosine and NECA, respectively, were consistent with actions at P1-purinoceptors. There was a trend towards an increase in the upper asymptote of the first phase of the NECA curve in the presence of increasing concentrations of 8-PT. The A1-purinoceptor selective antagonist, DPCPX, also blocked only the first phase of the NECA concentration-effect curve and produced a significant increase in the upper asymptote. The pA2 value (6.8) obtained was consistent with activation of A2-subtype P1-purinoceptors by the low concentrations of NECA. 3. There was no correlation between A1-purinoceptor affinity and the propensity to cause the increase in the upper asymptote of the first phase of the NECA concentration-effect curves amongst a series of 9-methyl adenine analogues, suggesting that the amplification was not due to inhibition of an underlying A1-purinoceptor-mediated contractile response.4. DPCPX (10 microM) produced a significant increase in the upper asymptote of the NECA concentration effect curve, but had no effect on isoprenaline curves whereas the phosphodiesterase inhibitor Ro20-1724 (30 microM) produced a significant increase in the upper asymptote of both NECA and isoprenaline concentration-effect curves. Therefore the amplification of the first phase responses by DPCPX did not appear to be due to phosphodiesterase inhibition.5. It was not possible to conclude whether second phase responses to adenosine and NECA were mediated by intracellular or extracellular sites of action. However, if intracellular sites of action were involved then adenosine did not apparently gain access by the Dip-sensitive uptake system.