OBJECTIVE: To examine by Holter electrocardiographic monitoring the effect of abruptly stopping nitrate treatment in patients with stable angina pectoris. PATIENTS: 12 men with confirmed ischaemic heart disease and stable exertional class 3 angina (Canadian). All had episodes of horizontal or down sloping ST segment depression during 24 hour electrocardiographic monitoring. All were nitrate responders. DESIGN: Each patient was given isosorbide dinitrate (10-30 mg four times a day) and placebo (four times a day) for three days in a randomised crossover trial. There was a washout period of 3-5 days between the two treatment periods. Holter monitoring was performed on the third day of isosorbide dinitrate and placebo administration and on the first day of their withdrawal. RESULTS: When treatment with isosorbide dinitrate was stopped there was a significant increase in the total number and duration of painless episodes of myocardial ischaemia. During placebo and isosorbide dinitrate administration 8 patients had episodes of painless myocardial ischaemia whereas after isosorbide dinitrate cessation they were recorded in all 12 patients. Episodes of silent myocardial ischaemia at rest appeared in 4 patients after isosorbide dinitrate withdrawal. CONCLUSION: Abrupt cessation of short-term continuous nitrate treatment in patients with severe angina may cause a rebound increase in myocardial ischaemia which is predominantly silent.