In this paper, we examine how students discuss and interpret data and whether these actions vary depending on the type of data they analyse. More specifically, we are interested in whether students perform differently when analysing first-hand data, which they collect themselves, compared with second-hand data provided to them. Our data analysis focused on two classrooms using two different curriculum units, chemistry in Grade 7 and biology in Grade 8, collected during the 2002/03 school year from a Mid-western urban middle school in the USA. We analysed classroom videotape associated with lessons in which students discussed first-hand and second-hand data both in small group settings and full class discussions. We found the two types of data offer different benefits and limitations, suggesting that both types of data are important for students to work with as they develop skills in scientific inquiry practices. We discuss the characteristics of classroom discussions around different data sources as well as implications for the design of curriculum materials, instructional environments, and student learning in science. (Contains 10 tables.)