A study was conducted to test the thesis that teaching word recognition skills in a manner compatible with the learner's auditory or visual modality preference would facilitate beginning reading instruction. A group of 62 students in two first grade classrooms was studied; one class stressed the presentation of words as whole units, the other stressed phonics. At the end of the academic year, correlation coefficients were computed between the difference scores that indicated modality preferences and scores in reading achievement. Results indicated the following: (1) no significant relationships were demonstrated between either the components or total modality preference scores and word recognition scores, regardless of teaching method; (2) those students in the phonics classroom with visual modality preference scored significantly higher on the word recognition test than did those with auditory preference; and (3) there was no significant difference in word recognition skills between classes when instruction stressed the whole word approach. (Author/DF)