Based on sampling from official census lists a sample of middle-aged women was obtained in five different age strata; 38, 46, 50, 54 and 60 years. Based on date of birth a sub-set was sampled for maximal exercise testing. The participation rate was 81% and 194 women underwent the test. In sixteen women the test was interrupted due to poor cooperation or clinical contra-indications against continued loading. 54% of the women considered the maximal load to be at least very straining according to a standardized scale for perceived exertion. The average maximal load decreased from 113 to 98 W with advancing age. In the highest age group only 16% exceeded 130 W compared to 43% in the youngest group. Maximum heart rate decreased with age but maximum respiratory frequency and perceived exertion did not. Systolic blood pressure 2 min after maximal work increased with age. Minnesota Code items 4:1-3 and 5:1-3 were demonstrated in 30% of the women, mainly in the older age strata. Other ECG-abnormalities were uncommon. Despite the widely differing incidence of myocardial infarction these ischaemic ECG-abnormalities were as common in women as in a comparative sample of men. This observation questions the traditional interpretation of the exercise induced ECG response in women.