This study investigated the stability and persistence of prospective elementary teachers' conceptual understanding of average speed. A questionnaire containing two questions about average speed was administered to the prospective teachers prior to their admission to an elementary science methods course (pretest) and at the end of the course (posttest). The pretest was administered to 84 prospective teachers, and the posttest was administered to 55 prospective teachers from the original sample of 84. A two-tier coding scheme was used to categorize responses as accurate and inaccurate definitions of average speed and to categorize the features of explanations of average speed. A chi-square test was carried out to test for statistical significance. Analysis of data indicated that prospective teachers used the conceptual understanding of speed to explain average speed. In addition to providing this explanation, prospective teachers used their conceptual understanding of average to explain average speed--that is, average speed as resulting from the calculation of the average of three or more recordings of an object's speed. The reasons for the two stable and persistent alternative conceptions could be that participants in this study were exposed to K-16 instruction that did not identify, label, represent, and distinguish the concept of average speed as a complex concept.