Important role of endogenous adenosine in the pathogenesis of hypoxic bradyarrhythmias was verified experimentally both in vivo and in vitro models of myocardial hypoxia in 180 guinea pigs and 40 rabbits by using enzyme differential spectrophotometry, synchronous quantitative analysis of correlated factors and adenosine antagonist. Results disclosed that endogenous adenosine in hypoxic myocardium increased markedly from a gram myocardial tissue 5.8 +/- 1.7 nmol of the control group to 27.5 +/- 2.0 nmol of 5% O2 inhalation group (P < 0.01). Sinus bradycardia and A-V block induced by injection of exogenous adenosine were closely akin to those caused by myocardial hypoxia both in ECG characteristics and dose-response curve. Both of their His electrograms showed prolongation of A-H interval but nochange of H-V interval, indicating the same locality of conduction block. Synchronous quantitative analysis of 3 related factors revealed a correlation coefficient of 0.99 (P < 0.01) between endogenous adenosine and bradyarrhythmias, denoting a subordinate instead of coordinate relation. Moreover, cause and effect relation could be shown between them from the result that aminophylline (50 mg, i.p.) could abolish completely II degrees-III degrees A-V block which would otherwise be bound to occur after inhalation of 5% O2. The present study is the first in vivo experimental verification of adenosine hypothesis throughout the world which not only can well explain those atropine-resistant cases, but also opens up the way to an entirely new therapeutic strategy for these kinds of bradyarrhythmias, i.e. instead of M-cholinoceptor blocker, adenosine antagonist (methylxanthines, e.g. aminophylline) should be prescribed.