A sample of 15 good and 15 poor first-grade readers, selected on the basis of the teacher's classification, performance on the Gates-MacGinitie Reading Test, Primary A, Form 1, and the Peabody Picture Vocabulary Test, was individually taught five nonsense syllables by each of four teaching modality procedures: visual, auditory, kinesthetic, and a combination of the three. The teaching procedure was based on the Mills Learning Methods Test and was carried out by the researcher in a laboratory situation. Twenty-four hours later, a test of retention was given. As expected, good readers took significantly fewer trials to master nonsense syllables and retained more nonsense syllables than the poor readers did. However, no single mode of learning resulted in significantly superior acquisition or retention of nonsense syllables for either good or poor readers as a group. Rather, modality preference appeared to be an individual matter. Limitations and educational implications of the study are given, and tables and references are included. (CL)