A population-based study of the incidence of craniosynostosis was conducted among residents of Olmsted County, Minnesota, born between 1 January 1976 and 31 December 1985. This study included only those with primary craniosynostosis who were less than 5 years of age at the time of diagnosis. Cases of craniosynostosis were classified as definite or probable. Definite cases had craniectomies or radiographic evidence confirming craniosynostosis, while patients for whom there was enough concern about craniosynostosis to order a skull radiograph were classified as probable cases. With these definitions, the incidence of craniosynostosis in Olmsted County was 3.1/10,000 births for definite cases (95% CI, 0.4-5.8) and 13.6/10,000 for all cases combined (95% CI, 7.9-19.3). These rates were compared to non-population-based minimal and maximal estimates of craniosynostosis frequency, 4/10,000 births and respectively. The rate observed in this study for definite cases, 3.1/10,000 births, is statistically consistent with the previous low estimate (observed/expected ratio 0.8, 95% CI, 0.2-1.8), and the rate observed for all cases in this study, 13.6, is consistent with the high estimate (observed/expected ratio 1.4, 95% CI, 0.8-2.1). Differences between definite and probable cases in this study are presented, along with comparisons with recent population-based data.