School personnel regularly conduct systematic direct observation (SDO) as part of educational evaluation and intervention progress monitoring. Despite the regularity with which SDO is utilized, little research has been devoted to the identification of efficient and effective training procedures. Thirty-two undergraduate students in education participated in the current study and were randomly assigned to one of two methods of feedback as a means of promoting accurate SDO: post hoc feedback and real-time feedback. Participants in each group completed three observations, one pre-training, one training, and one post-training. The dependent variables assessed were changes in interobserver agreement (IOA) with a criterion and overall accuracy of the observation. Results did not reveal significant differences in IOA between the two groups. A significant interaction was found between group assignment and measurement occasion for observation accuracy. Overall, results suggest that real-time feedback did not confer added benefit.