A total of 117 subjects were randomly assigned to a group receiving instructional objectives, or not doing so, and to a logical or random instructional sequence. Performance measures, as well as test and state anxiety scores were obtained. Regression analysis indicated that, as expected, objectives had no effect, whereas the logical sequence reduced program errors and increased achievement. Test anxiety was related to program errors, but not to achievement. Expected interactions among objectives, sequence, and anxiety were not significant.The use of behavioral objectives in instruction has been a topic of much concern to educators in the past decade. The view that objectives enhance achievement has been actively investigated and challenged in some instances. Similarly, the sequence with which instructional material should be presented has also received considerable attention, and often led to the surprising finding that random instructional sequences yielded achievement which did not differ from that gained with carefully prepared sequences. The first purpose of this experiment was to study these two questions, and their interaction. A further purpose was to investigate