We studied the effect of current smoking habits in the period immediately before admission to hospital in 90 consecutive patients presenting with chest pain, 50 of whom were shown to have myocardial infarction. Urine cotinine/creatinine (cot/creat) ratio measured within 4 hours of admission was used as an objective marker of cigarette smoking in the preceding 18 hours. Fifty-seven patients had urine cot/creat ratios suggesting recent smoking, although four of these denied smoking. Patients with myocardial infarction had higher median cot/creat ratios (3.31 micrograms/mg, range 0-17.8) compared with patients with non-infarct chest pain (0.5 microgram/mg, range 0-37.2). Sixteen patients with cardiac rhythm disturbances following infarction had significantly higher cot/creat ratios than the 34 infarct patients without this complication (median and range 8.34 micrograms/mg; 0-17.8 V. 1.87 micrograms/mg; 0-16.4, P < 0.01). Tobacco use in the 24 hours before myocardial infarction may predispose to cardiac rhythm disturbance, irrespective of infarct size.