Since therapeutic decisions in patients with angina pectoris are usually based on the reported frequency of exertional and rest pain the relations between the historical frequency of chest pain and objective evidence of myocardial ischaemia during normal daily activity were investigated in 100 patients by 48 hour ambulatory ST segment monitoring. Of these 100 consecutive patients with chest pain, 91 had typical pain and nine some atypical features. Twenty six patients had normal coronary arteries and 52 of the 74 with significant coronary disease had ambulatory ST segment changes. There was no relation between the frequency of reported exertional or rest pain and (a) the severity of coronary artery disease or (b) the frequency of daytime or nocturnal ST segment changes. Twelve patients had nocturnal ST segment changes but only four complained of nocturnal angina. Most patients had both painful and painless episodes of ST segment changes, but a substantial number had either painless or painful episodes only. These differences were not related to the severity of coronary artery disease. Chest pain after the onset of ST segment change was perceived with wide interpatient and intrapatient variability. Thus the frequency of pain is a poor indicator of the frequency of significant cardiac ischaemia. Individual differences in the perception of pain may be more important.