Beschin and Robertson (1997) devised a simple clinical test of left personal neglect, which characterises personal grooming behaviour according to the proportion of the total activity that is directed to the left side of the body. Although this test proved highly reliable, and more sensitive than prior diagnostic techniques, its formulation may yet be improved. The present paper reports a reanalysis of Beschin and Robertson's (1997) data, using additional control subjects, and a formula which characterises personal neglect as a lateral bias of behaviour rather than as a lateralised deficit. It is shown that this formula greatly enhances the test's sensitivity to the behavioural abnormalities of brain damaged patients, and it is recommended that this modification be adopted for the future diagnosis of personal neglect.