Since one way to study a non-reading primary child's phonics knowledge is to examine his/her invented spelling, a researcher's quandary led to a quasi-experimental design study, employed to answer three questions: (1) Do primary non-readers possess phonics knowledge? (2) Do primary non-readers possess phonemic awareness? and (3) Do primary non-readers spell at a high developmental level? If many reports and authorities are accurate, findings should be that non-readers lack these things, hence their inability to read. Subjects were 50 primary students from a small rural city in the southeastern United States, identified by their teachers as non-readers. As part of their routine, these children were instructed from highly phonetic basal readers and experienced daily drills from a commercially prepared phonics program. Following a screening, 26 students were identified as reading less than 80% of the words from a passage in their primer and were classified as non-readers. After some were excluded for various reasons, the final sample was comprised of 21 children. The children were tested individually, and findings revealed that these students were phonetically knowledgeable, possessed phonemic awareness, and could spell at a level which includes the recognition of letter name vowels, consonants, and some short vowels. What was apparent was that they had been taught to use only one cueing system--phonics. If children actively construct knowledge through assimilation and re-inventions, then the approach to the teaching of reading must be to provide support for every child in his/her development of all three cueing systems. (Contains 3 figures and 17 references.) (NKA)