It has been suggested that central nervous system (CNS) neuroexcitation plays an important role in digitalis-induced cardiac arrhythmias. To elucidate further the role of the CNS in digitalis-induced arrhythmias, the inotropic and toxic effects of a highly polar semisynthetic cardiac glycoside, 3beta-O-(4 amino-4,6 dideoxy-beta-D-galactopyranosyl)-digitoxigenin (ASI-222) were compared to those of digoxin and correlated with plasma and cerebrospinal fluid (CSF) concentrations of each drug. Thirteen dogs anesthetized with sodium pentobarbital were given repeated intravenous doses of digoxin or ASI-222. Ventricular tachycardia was elicited at a mean dose of digoxin of 0.12 mg/kg, compared with 0.09 mg/kg for ASI-222 (not significant). Terminal ventricular fibrillation occurred after 0.18 mg/kg of digoxin, a value significantly larger than the ASI-222 dose (0.14 mg/kg, P less than 0.05) required to produce lethal arrhythmias. Digoxin produced a 21% increase in LV dP/dt at a plasma digoxin concentration of 20.0 +/- 2 ng/ml (mean +/- SEM) 30 minutes after 0.05 mg/kg; the CSF digoxin concentration at this time averaged 0.7 +/- 0.1 ng/ml. At death, the plasma digoxin concentration was 88 +/- 16 ng/ml and CSF digoxin concentration was 5.7 +/- 1.6 ng/ml. ASI-222 produced a similar 25% increase in LV dP/dt 30 minutes after administration of 0.05 mg/kg, with a plasma concentration of 18 +/- 2 ng/ml as determined by a newly developed radioimmunoassay. The plasma ASI-222 concentration at death, 95 +/- 18 ng/ml, was similar to that of digoxin. However, CSF samples at 30 minutes and at death showed no detectable levels of ASI-222. Thus, despite similar inotropic and toxic responses and similar plasma drug concentrations compared to digoxin, ASI-222 was demonstrated to have limited if any access to the CNS as judged by CSF concentrations. These findings suggest that direct CNS stimulation does not play a primary part in the genesis of digitalis-induced cardiac arrhythmias in this experimental model, or that CNS effects are mediated by an area or areas lacking an effective blood-brain barrier.