Early childhood teachers are increasingly encouraged to support children's development of letter-sound abilities. Assessment of letter-sound knowledge is key in planning for effective instruction, yet the letter-sound knowledge assessments currently available and suitable for preschool-age children demonstrate significant limitations. The purpose of the current study was to use item response theory to create short-form letter-sound assessments that are psychometrically sound, quick and easy to administer, and appropriate for researcher and practitioner use. Letter-sound knowledge data from 940 children attending preschool centers were analyzed. Item response theory analyses indicated that a two-parameter logistic model best fit the data. Further analyses yielded a series of parallel six- and eight-letter forms with adequate test information, reliability, and theta recovery. Implications for assessment and instruction are discussed. (author abstract)