Background: In this study, 2 main hypotheses have been put forward to explain the variation in childhood intelligence or school performance by season of birth. In the first hypothesis, it is suggested that it is due to school policy concerning school entry, whereas the second suggests that a seasonally patterned exposure such as temperature, maternal nutrition, or infection during critical periods of brain development have a lasting effect on intelligence. Aims: To determine whether childhood performance on tests of different domains of intelligence is patterned by season of birth and to examine possible mechanisms for any associations. Sample: 12,150 individuals born in Aberdeen, Scotland between 1950 and 1956. Methods: Birth cohort study in which the variation in different domains of childhood intelligence measured at ages 7, 9, and 11 by season of birth were examined. Results: Reading ability at age 9 and arithmetic ability at age 11 varied by season of birth, with lowest scores among those born in autumn or early winter (September-December) and highest scores among those born in later winter or spring (February-April); p = 0.002 for joint sine-cosine functions for reading ability at age 9 and p = 0.05 for sine-cosine function for arithmetic ability at age 11. The child's perception and understanding of pictorial differences at age 7, verbal reasoning at 11, and English language ability at 11 did not vary by season of birth. Age at starting primary school and age relative to class peers were both associated with the different measurements of childhood intelligence and both attenuated the association between month of birth and reading ability at age 9 and arithmetic ability at age 11 towards the null. Both adjusted and unadjusted differences in reading ability at age 9 and arithmetic ability at age 11 between those born from September to December compared with other times of the year were less than 0.1 of a standard deviation of the test scores. Ambient temperature around the time of conception, during gestation, and around the time of birth did not affect intelligence. Conclusion: Any variation in mean childhood intelligence by season of birth is weak and largely explained by age at school entry and age relative to class peers.