Background: In the last decade, there has been a growing interest, both in the UK and abroad, in developing early screening tests for dyslexia that can be used with very young children. In addition to measures of literacy achievement, such screening tests typically aim to identify underlying difficulties, such as phonological deficits, that might hinder a child's educational progress. The Dyslexia Early Screening Test (DEST; Nicolson & Fawcett, 1996) is an example of such a test that combines attainment and diagnostic indicators. Aim: The study reported assessed the ability of the DEST to predict future literacy skills in contrast with the prediction afforded by school-based measures, such as letter knowledge. Sample: Participants were 45 boys attending a reception class, with a mean age of 4.87 years at the start of the study and 6.63 years at the end. Methods: Measures of literacy skills, phonological awareness, verbal memory, motor skill, and auditory processing were assessed using the DEST as the initial screening tool at Phase 1. Measures of letter knowledge, non-word reading, and rhyme judgment were taken at Phase 2. Phase 3 measures, 14 months after the start of the study, comprised single-word reading and spelling. At Phase 4, some 22 months after the beginning of the study, measures of reading and spelling ability were assessed again. Results: Individual subtests of the DEST were more predictive of later literacy skills than the global screening test's score (the 'at risk quotient'). Better predictors were the DEST subtest of sound order and rapid automatized naming, together with the school attainment measure of letter knowledge. Conclusions: Although some DEST subtests did offer predictors of future literacy skills, school-based measures of letter knowledge may be equally valid as assessment measures. Additionally, the results question the usefulness of combining measures to form an 'at risk' index of future literacy difficulties, particularly in the age range assessed.