In a modification of conventional walking tests in respiratory disease patients were asked to walk successive 50 m lengths of a corridor and the time taken over the third 100 m was taken as the walking pace. This pace was highly correlated (r = 0.994) with the 12-minute distance and was highly reproducible (r = 0.98).

SIR,-We read with interest the short report by Dr R Butland and others (29 May, p 1607) on walking tests in respiratory disease. We agree that while a 12-minute walk is a useful measure of exercise tolerance it is time consuming to perform. They attempted to shorten the test by reducing the specified time in which the patient walks from 12 to six or even two minutes. This, however, produces difficulties since, as they and others" have observed, patients usually start walking at a faster pace before settling to a remarkably constant speed. Shortening the walking time thus tends to measure the initial spurt rather than the true exercise tolerance. A much better method is to determine the patient's walking pace over a constant distance; we have found 100 m to be the optimum distance (two lengths of a 50 m corridor). Patients are instructed as for a 12-minute walk and the time taken for each 50 m is recorded. Most patients rapidly reach a constant pace, as shown by almost identical 50 m times. Time taken to walk the next 100 m is then recorded, and if required the 12-minute walking distance may be calculated. Initial results in 13 patients show a high degree of correlation between actual and calculated walking times. We are hopeful that the 100-m walk will prove the simplest, quickest, and most accurate method of producing an index of exercise tolerance in respiratory disease. ALYN MORICE T SMITHIES