Used 4 experimental tasks to test the effects on procedural performance of providing special instructions on logical-tree construction and use and of limiting vs not limiting the time available for studying the task instructions. Ss were 20 male undergraduates. Results indicate that performance accuracy was statistically better when either one or both logical-tree instruction and practice was provided and the task instruction study time was limited, than when Ss were permitted to study the task instruction for as long as they chose and in whatever way they chose. It is concluded that quite simple procedures for familiarizing Ss with logical-tree operations can improve performance on procedural tasks. It also appears that placing a limit on the time available for study of instructions can be better than permitting unlimited time.