High school students took part in a series of pilot studies on the variables involved in step size and sequencing in programed instruction. Programed learning materials of the University of Illinois Committee on School Mathematics (UICSM) were varied as to covert and overt trial sequences, use of teachers, type of prior practice, order of frames, and size of step. Learning time was not significantly different when sequences of overt and covert trials were varied. When teachers lectured before or after students used programed materials, no significant difference in performance was found between the two conditions. Type of prior practice was varied three ways: stimulus-practice, mediator-practice, and response-practice, all of which were equal in their effect on performance. The order of appearance of items in a program was found to make no difference in learning. In a final study comparing large and small steps, the large step program was devised by omitting non-essential frames from the original program. A t-test indicated a significant saving of time for subjects using large step materials. It was concluded that studies which further investigate these variables should be based on larger samples. (LS)