Two groups of sighted pre-school children were taught to name six braille letters: one group received phoneme instruction and the other grapheme instruction. Ten boys and ten girls (average age 4:5 years) participated. There was a statistically significant advantage for the phoneme group (Experiment 1). In a repeated measures design, 16 sighted primary-school children (8 boys and 8 girls), with an average age of 10:9 years, were first taught to name 10 braille letters as phonemes, and another 10 braille letters as graphemes (Experiment 2). Then the same children were taught to name 10 braille words as onset-rime and another 10 braille words as whole words (Experiment 3). There was a statistically significant advantage for both phoneme instruction and onset-rime instruction. (Contains 7 tables and 4 figures.)