Aims: School lunches are provided free in Sweden, although some children choose not to eat school lunch. The aim of this study was to analyse Swedish children's total energy and nutrient intakes on weekdays by the frequency of school lunch consumption and to analyse energy and nutrient intakes from school lunches by sex. Factors associated with children's school lunch habits were also studied. Methods: Children in grades 2 and 5 (n=1905) completed a food diary (school lunch data available for 1840 children) and the mean energy and nutrient intakes per day and per school lunch were calculated. The children also completed questions on the frequency of school lunch consumption and school lunch habits. Logistic regression was used to assess factors associated with school lunch habits. Results: Children who reported eating school lunch every day had significantly higher energy and absolute nutrient intakes than children reporting eating school lunch less than five times a week, but not standardized for energy. Boys had significantly higher energy and absolute nutrient intakes from school lunches than girls, but not standardized for energy. Younger children and children who liked school lunches had higher odds of eating school lunch every day. Children in grade 5, those with a foreign background and those disliking school lunches had higher odds of omitting the main lunch component. Conclusions: Regular school lunch consumption was associated with a higher total intake for most nutrients, but not a better nutrient density. School lunch habits were associated with age, ethnic background and liking school lunches.