AIM: The purpose of this study was to assess the relation of perinatal risk factors with later development of specific language impairment (SLI). METHOD: In a case-control study, 179 children attending special needs schools for SLI were matched with non-affected children attending mainstream schools. Both groups consisted of 134 males and 45 females (age range 4-13y; mean age 9y, SD 2y 4mo). Data on duration of pregnancy, birthweight, delivery complications, birth characteristics, and Apgar scores were collected from the Preventive Child Health Care files of the Municipal Health Service. RESULTS: The gestational age of the children with SLI (mean 39.6wks, SD 0.1wk) and for the comparison group (mean 39.4wks, SD 0.1wk) and the birthweight of children with SLI (mean 3330.4g; SD 41.4g) and for the comparison group (mean 3388.1g; SD 39.8g) were not statistically different; neither were other pregnancy and birth characteristics, with the exception of the Apgar scores (effect of group for Apgar score after 1min p=0.045; after 5min p=0.001). The difference in Apgar scores was larger for females than for males (effect of group x sex for Apgar score after 1min p=0.049; after 5min p=0.043). INTERPRETATION: The relation between Apgar scores and SLI together with the influence of sex may be meaningful for predicting modelling and for understanding the causal pathway for SLI.