The effects of beta-adrenoceptor antagonists on salivary secretion have been extensively studied in animals but not in man. The aim here was to compare salivary flow rate and composition in man during 1 week of treatment with a non-selective (propranolol 80 mg b.i.d.) and a beta 1-selective (atenolol 50 mg o.d.) antagonist with that of placebo. The randomized, double-blind, cross-over ("Latin square") design was used and 42 healthy male volunteers were recruited to the study. The treatment periods were separated by a wash-out period of 2 weeks. Whole saliva was sampled on day 0 (before) and on day 7 during each treatment. The plasma concentration of propranolol and atenolol was determined from blood samples obtained on day 7. Resting saliva was assessed for flow rate, amylase activity and concentration of total protein, hexosamine and sialic acid. Stimulated saliva was assessed for flow rate, pH, buffer pH, amylase activity and concentration of total protein, Ca2+, Mg2+, Na+, K+, Cl- and PO4(2-). In resting as well as stimulated whole saliva both the total protein concentration and the amylase activity were significantly decreased during the active treatment periods (p less than 0.05-p less than 0.001). At lunchtime during atenolol treatment the hexosamine/total protein and the sialic acid/total protein ratios were significantly increased (p less than 0.05-p less than 0.01), suggesting a possible effect on protein synthesis. In addition, the concentrations of Ca2+, PO4(2-), Cl- and Mg2+ were significantly altered during the active treatment periods (p less than 0.05-p less than 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)