It is well documented that traditional introductory physics courses often fail to teach our students the basics. Indeed, these same courses often teach students things we don't want. Beyond the usual focus on pure content, there are extensive sets of attitudes and beliefs about science that we teach our students. Some of these messages are beneficial (e.g., that science is a coherent representation of the world) while others are detrimental (e.g., the notion that women cannot be strong physicists). While decades of physics education research have reformed classroom practices to improve conceptual mastery, these same practices often fail to improve student attitudes and beliefs about learning physics. This workshop presentation introduces the development and structure of a new instrument designed to probe students' beliefs about science and learning science, the Colorado Learning About Science Survey (CLASS) [1]. In addition to exploring the various dimensions of student beliefs that this instrument probes, it examines correlations between student beliefs and class performance, students' understanding of expert beliefs, and gender differences. It also covers issues of class environments and practices that support productive beliefs.