It is generally believed that students should use multiple representations in solving certain physics problems. In this study, we interview expert and novice physicists as they solve two types of multiple representations problems: those in which multiple representations are provided for them, and those in which the students must construct their own representations. We analyze in detail the types of representations subjects use and the order and manner in which they are used. Somewhat surprisingly, both experts and novices make significant use of multiple representations. Some differences emerge: Expert use of multiple representations is more dense in time, and novices tend to move between the available representations more often. In addition, we find that an examination of multiple representation use alone is inadequate to fully characterize a problem-solving episode; one must also consider the purpose behind the use of the available representations.