Twenty-four nonsmoking male volunteers took 50 mg atenolol or 10 mg betaxolol orally once a day for 9 days in a two-period, four-sequence, randomized, crossover study. Plasma concentrations reached steady state after day 5. Percent fluctuation in plasma concentration defined as (Cmax–Cmin)/Cavg(% fluctuation 1) was 97% on day 9 for betaxolol and 343% for atenolol; thus atenolol fluctuation was more than threefold that of betaxolol. A 10-fold difference in plasma level fluctuation was observed when fluctuation was defined as (Cmax–Cmin)/Cmin(% fluctuation 2). The intersubject variances for % fluctuation 1 and % fluctuation 2 were 4.1 and 85.5 times greater for atenolol than for betaxolol; these differences were marginally statistically significant for % fluctuation 1 and significant for % fluctuation 2. The intra-subject variabilities for area under the curve and plasma level fluctuations were statistically greater for atenolol than for betaxolol. Atenolol intrasubject variances were 25 and 271 times greater than for betaxolol for % fluctuation 1 and % fluctuation 2, respectively. Thus, betaxolol exhibited less fluctuation in plasma levels with substantially less intersubject and intrasubject variability. These factors would be expected to provide a more consistent therapeutic response and more dependable dosage adjustment.