A study compared the effects of a modified whole language approach with those obtained from a pure whole language approach on the decoding abilities of 20 kindergartners, assessed using Darrell Morris's Early Reading Screening Instrument. Subjects were selected from two similarly composed schools located in Albemarle County, Virginia. Ten students were taught in a pure whole language classroom, and the other 10 students received instruction using a modified whole language approach with about 20 minutes per day devoted to focused, intensive, and systematic instruction in phonics. Results indicated that the whole language students receiving a daily supplement of explicit phonics instruction demonstrated greater alphabet knowledge, phonological awareness, word recognition in context, and word recognition in isolation than students receiving pure whole language instruction. Findings suggest that the modified whole language approach has a greater effect on student decoding automaticity than the pure whole language approach with which it was compared. (Contains 19 references and 4 tables of data. An assessment instrument is attached.) (RS)