Abstract : The problem of quantifying a simple, often-used clinical ataxia test (tandem-walking) for inclusion in an ataxia test battery was overcome by adopting the method of counting the number of heel-to-toe steps (1 to 10) a person can take without sidestepping with eyes closed and arms folded against chest. Standardization under these rigid conditions was based on testing 287 normal men and 100 normal women. Validation was based on testing 22 individuals having labyrinthine defects of varying severity and origin. This newly quantified test was found to be as valid an indicator of vestibular ataxia as other subtests of a multidimensional quantitative ataxia test battery with which it was compared, and it has the advantages that, it appears to be free of age influences, and nearly all the normal and none of the abnormal individuals obtained a perfect score.