Rasch analysis was used to examine the validity of the Children's Fear Survey Schedule – Dental Subscale (CFSS-DS). Data were collected from parents of children in regular dental care (n = 240) and from parents of children referred to a specialized pediatric dental clinic (n = 200). The Rasch model is a probabilistic model in which estimations are based on the relationship between person ability and item difficulty, and the present study investigated the validity of the CFSS-DS through analysis of person and item fit, dimensionality, ordering of thresholds, local dependency, and differential item functioning. The results show multidimensionality of the original scale, and an adjusted version consisting of six items with good fit to the model is suggested. The adjusted scale makes interval-level analysis possible, and time efficiency and elimination of previously criticized items argue in favor of the adjusted scale.