Sister chromatid exchange (SCE) and micronuclei (MN) were studied in peripheral human lymphocytes, treated in vivo with atenolol, azathioprine (AZA) and Chlorpropamide (CPA), and from patients on long term therapy with these drugs. The clinical study consisted of samples from 18 atenolol, 8 AZA and 9 CPA treated patients, and 35 age and sex matched controls. The CPA patients' controls were matched for disease (diabetes). Additional controls monitored variation between culture time. 20 metaphases were scored for SCEs and 2000 cells for MN per person. Statistical comparison of treated and untreated groups were with two-sample t-tests (SCE data) and 2x2 contingency χ² tests (MN data). There were two significant results. The atenolol treated had more MN (p< 0.01) and the AZA treated more SCEs (p< 0.02) than their control groups. The CPA treated and control groups did not differ. In the <i>in vitro</i> study concentrations used were, Atenolol 1.3, 2.7, 6.7 and 13.3ug/ml, AZA 5, 25, 50 and 100ug/ml and CPA and Diabinese (DIA) (pharmaceutical preparation of CPA) 250, 500 and 1000ug/ml. All culture lengths were 72 hours after mitotic stimulation (0 hours). Several exposure regions were used, they were: With metabolising enzymes (S9), all drugs for 6 hours at 48 hours. Without S9, Atenolol for 72 hours, AZA for 24 hours (at times - 24 hours and 48 hours), CPA and DIA for 24 hours at 48 hours. 20 metaphases were scored for SCEs and 4000 cells for MN. Dosed cultures were compared statistically with the control cultures using analyses of variance (SCE data) and linear regression analyses (MN). The significant results were; AZA, increases in SCEs in all experiments, increases in MN with S9 and without S9 (experiment b); CPA increase in MN with S9, decrease in MN without S9. No effects were found with atenolol or DIA treatment.