Cumulative concentration-response curves to isoproterenol were constructed in 74 preparations of human airways (group A) contracted with acetylcholine (80% of maximal contraction). In 52 bronchi (group A1), the maximal relaxation to isoproterenol represented at least 70% of the acetylcholine contraction (average 98% +/- 3) and the mean concentration (+/- SD) causing 50% of the relaxation (EC50) was 6.0 X 10(-8)M +/- 0.8. Resistance to isoproterenol (significant decrease of the maximal relaxation and right shift of the concentration-response curve) was provoked in 25 preparations by incubating them with isoproterenol (concentration 100 times higher than the individual EC50) for 30 min. The response to isoproterenol remained stable over the same time interval in eight control preparations. Incubation with isoproterenol did not modify the relaxing response to theophylline (n = 8). Indomethacin had no effect on the resistance provoked by isoproterenol incubation (n = 11). Incomplete relaxation to isoproterenol (maximal relaxation averaging 47% +/- 2 of the acetylcholine contraction) was observed in 22 preparations (group A2; mean EC50 = 1 X 10(-6)M +/- 0.2, significantly different [p less than 0.001] from group A1). Such preparations could be completely relaxed by theophylline (n = 7). For the whole group A, a significant negative correlation was found between isoproterenol EC50 and the magnitude of the maximal relaxation to isoproterenol (expressed as percent of acetylcholine contraction). There was no significant correlation between acetylcholine EC50 and isoproterenol maximum relaxation. Spontaneous resistance to isoproterenol was neither related to the technique of isoproterenol administration (cumulative versus noncumulative; n = 11) nor the magnitude of acetylcholine contraction (70% and 95% of maximal contraction; 14 bronchi, group B).