The gender similarities hypothesis by J. S. Hyde ( 2005 Hyde, J.S. (2005). The gender similarities hypothesis. American Psychologist, 60, 581–592. [Crossref], [PubMed], [Web of Science ®] [Google Scholar] ), based on large-scale reviews of studies, concludes that boys and girls are more alike than different on most psychological variables, including academic skills such as reading and math (J. S. Hyde, 2005 Hyde, J.S. (2005). The gender similarities hypothesis. American Psychologist, 60, 581–592. [Crossref], [PubMed], [Web of Science ®] [Google Scholar] ). Writing is an academic skill that may be an exception. The authors investigated gender differences in academic achievement using a large, nationally stratified sample of children and adolescents ranging from ages 7–19 years (N = 2,027). Achievement data were from the conormed sample for the Kaufman intelligence and achievement tests. Multiple-indicator, multiple-cause, and multigroup mean and covariance structure models were used to test for mean differences. Girls had higher latent reading ability and higher scores on a test of math computation, but the effect sizes were consistent with the gender similarities hypothesis. Conversely, girls scored higher on spelling and written expression, with effect sizes inconsistent with the gender similarities hypothesis. The findings remained the same after controlling for cognitive ability. Girls outperform boys on tasks of writing.