Fifty-two students from two universities who had demonstrated mastery of the use of behavioral objectives were randomly assigned either ten knowledge level objectives (K), ten above knowledge level objectives (A), or to a non-objectives control group (C). All were tested on the same prose material using knowledge and above knowledge level items correlated with each objective. Results of the 3 X 2 analysis of variance indicated significant main and interaction effects beyond the 0.05 level on the knowledge level test. A Scheffe' analysis of the main effect revealed that the K group out performed the A group and the Scheffe' analysis of interaction showed that one A group performed significantly worse than one K group. A questionnaire revealed no significant differences in learning strategies employed. (Author)