This paper provides normative data for Australian school children on a modified version of the Castles Word/Non-Word Tests (Castles, 1993). The tests were designed to isolate the lexical and nonlexical reading procedures. Data were collected from 298 school children in Perth and combined with data provided by Coltheart and Leahy (1996) for 420 school children in Sydney. Norms for the Sydney sample have been published previously (Coltheart & Leahy, 1996). Norms for the combined sample are reported in 12-month age bands from 7 to 12 years, in the form of normalised standard scores. Issues surrounding subtyping research in dyslexia are reviewed, and a way to subclassify research samples using the provided norms is outlined and evaluated.