Background: Several studies in recent years have indicated a link between socio-economic status (SES) of families and children's language development, including studies that have measured children's language through formal standardized test procedures. High numbers of children with low performance have been found in lower socio-economic groups in some studies. This has proved a cause for concern for both clinicians and educationalists. Aims: To investigate the relationship between maternal education and postcode-related indicators of SES, and children's performance on the New Reynell Developmental Scales (NRDLS). Methods & Procedures: Participants were 1266 children aged between [2.0] and [7.06] years who were recruited for the standardization of a new assessment procedure (NRDLS). Children were divided into four groups reflecting years of maternal education, and five groups reflecting SES Index of Multiple Deprivation (IMD) quintiles for the location of participating schools and nurseries. Groups were compared using analysis of covariance (ANCOVA), with age as a covariate, in order to identify which might be affected by the two SES variables. Where relationships were found between SES and performance on the scales, individual children's standard scores were looked at to determine numbers potentially at risk for language delay. Outcomes & Results: An effect of years of maternal education on performance was found such that children whose mothers had minimum years performed less well than other children in the study, this effect being stronger for younger children. Children attending schools or nurseries in IMD quintile 1 areas performed less well in language production. Higher than expected numbers with language delay were found for younger children whose mothers had minimum years of education, and for children in quintile 1 schools and nurseries; however, numbers were not as high as noted in some other studies. Conclusions & Implications: Characteristics of the participant sample and measures used for language and SES may explain these results and are important considerations when interpreting results of studies or developing policies for intervention. The usefulness of commonly used categories of language delay is questioned. (Contains 5 tables and 5 figures.)