One hundred and three patients with a history of ventricular arrhythmia were screened by baseline 24-h Holter monitoring for a minimal mean number of 30 premature ventricular beats (PVB) per hour. One hundred and one patients were subsequently enrolled in a double-blind, cross-over trial, including four 4-day periods with administration of placebo (PL) or nicainoprol (N.) in doses of 200, 400, 600 mg three times daily in randomized order. Antiarrhythmic efficacy could be evaluated in 61 patients who met all requirements of the protocol, including a complete sequence of valid Holter recordings (greater than 18 h each). Total PVBs were significantly reduced by the 400-mg dose (65%) and the 600-mg dose (71%) as compared to baseline. Two hundred mg N. and PL led to a non-significant reduction of PVBs by 41% and 24%, respectively. An individual response, assumed where there was simultaneous reduction of total PVBs greater than 75%, of ventricular pairs greater than 90% and elimination of ventricular runs, was found in 6 (10%), 10 (16%), 21 (34%), and 22 (36%) patients with PL, 200 mg N., 400 mg N. and 600 mg N., respectively. Four hundred and 600 mg N, significantly prolonged the PR interval (0.17 s and 0.18 s; median values) as compared with baseline (0.16 s), while the QRS duration was significantly increased only by the 600-mg dose (from 0.10 to 0.11 s). Heart rate, mean blood pressure, QT and JTc intervals were not significantly influenced by any dose of N.(ABSTRACT TRUNCATED AT 250 WORDS)