Abstract
The gait pattern of five amputees was recorded by visual observation and also by a quantitative measurement system. The results were compared with the gait pattern predicted from biomechanical analysis. Visual observation was found to be an unreliable clinical skill. The observers recorded only 22.2% of the predicted gait deviations and were unable to comment on 15.6% of all the required observations. The gait parameters which were difficult to assess visually, namely step length and step time, are considered by most workers to be of particular value as diagnostic clinical indicators in gait assessment. The measurement system used was accurate in recording gait deviations and picked up 3.4 times as many deviations as visual observation. This report demonstrates the inadequacy of visual observation as a diagnostic method and emphasises the need for measurements.