The influence of vagal and sympathetic efferent activity on sinus arrhythmia in man has been studied in six healthy subjects by administration of hyoscine butylbromide and/or various beta-adrenoceptor blocking drugs using a microcomputer-linked electrocardiogram system. Sinus arrhythmia was quantitated as the s.d. of the R-R interval. Sinus arrhythmia was almost abolished by hyoscine butylbromide irrespective of the absence or presence and nature of the beta-adrenoceptor blocking drug. Atenolol and metoprolol alone prolonged the mean R-R interval and increased sinus arrhythmia. Oxprenolol, a drug with modest partial agonist or intrinsic sympathomimetic activity (ISA), prolonged the mean R-R interval to a lesser extent but had no effect on sinus arrhythmia. Xamoterol, which has high ISA, shortened the mean R-R interval but had no effect on sinus arrhythmia. These data yielded a non-linear relationship between sinus arrhythmia and mean R-R interval. Exaggerated sinus arrhythmia appears to accompany beta-adrenoceptor blockade only in the absence of ISA when bradycardia ensues. These findings are consistent with the hypothesis that the exaggeration in sinus arrhythmia is due to a central vagotonic effect secondary to the action of the drugs in the periphery. Changes in R-R interval induced by the adrenoceptor blocking drugs were altered to some extent by vagal blockade. This observation is consistent with the hypothesis that changes in heart rate induced by such drugs are determined in part by a change in vagal tone.