Two third grade classes in a private school in North Carolina were given keyboarding instruction using Sunburst's "Type To Learn." So that the effects of color-coding could be examined, one class was given standard keyboards to use (control group), while the other was given keyboards that were color-coded according to proper finger placement (treatment group). Neither class had any previous keyboarding instruction. Before the study, students in both classes typed a paragraph using "Typing Tutor IV." This software measured their speed and accuracy; the mean speed and mean accuracy scores of each were used to determine comparability. During the study, students met twice a week for a 30-minute class. Scores were averaged so that a mean score for speed and a mean score for accuracy from each class could be compared using two tail t tests. Results indicated that implementing keyboarding instruction that uses color-coded keyboards with these students did not appear to improve either keyboarding skills or accuracy. (Contains 34 references.) (Author/AEF)