Returns of scheduled school entry examinations, from 838 primary school children in West Lambeth Health Authority, were analysed to assess the possible association of a child's home language background, age, sex, behaviour, and examiner with the results of the developmental examination. To allow for variation in social characteristics between participating schools, a ward index of deprivation based on the 1981 census was assigned to all the children in the schools of each ward. Variables representing development were most consistently associated with the child's behaviour during the examination and with the examiner. A child's home language background was associated only with the auditory memory test. The child's ability to concentrate during the test situation was closely related to his performance. With respect to the examiners, it is suggested that community health services need to consider schemes of ongoing in-service training and careful calibration of the tests.