Walking tests, frequently used to document effects of treatment on exercise capacity, have never been standardised. We studied the effects of encouragement on walking test performance in a randomised study that controlled for the nature of the underlying disease, time of day, and order effects. We randomised 43 patients with chronic airflow limitation or chronic heart failure or both to receive or not receive encouragement as they performed serial two and six minute walks every fortnight for 10 weeks. Simple encouragement improved performance (p less than 0.02 for the six minute walk), and the magnitude of the effect was similar to that reported for patients in studies purporting to show beneficial effects of therapeutic manoeuvres. Age and test repetition also affected performance. These results demonstrate the need for careful standardisation of the performance of walking tests, and suggest caution in interpreting studies in which standardisation is not a major feature of the study design.