The remediation of gait problems is a key feature of neurological physiotherapy. The ten-metre walking test is a frequently used outcome measure in this clinical area. Despite its simplicity and apparent popularity, the ten-metre walking test does not have an established operational procedure. This article proposes a procedure, together with a brief account of its origins and features. Using the procedure, acceptable levels of inter-rater reliability were found, with a 95% CI for agreement of −0.38 to +0.38 seconds for normal young subjects and −0.36 to +0.49 seconds for traumatic brain-injured subjects. The ten-metre test results for two groups of head-injured patients are also reported. A focus of the discussion is how the test might best be used in clinical practice, as well as indications for future research still needed. An overall aim of the paper is to commend the continued considered use of the test by physiotherapists working in neurology.