Informed by decades of research and standards-based policies, there has been a growing demand for high-quality teaching and learning in mathematics and science classrooms. Achieving these ambitious goals will not be easy; students' opportunities for learning as shaped by the tasks they are assigned will matter the most. The purpose of this article is to revisit theory and research on tasks, a construct introduced by Walter Doyle nearly 40 years ago. The authors discuss how this construct has been used and expanded in research since then, argue for its applicability to contemporary challenges facing schools and classrooms today, and provide suggestions for future research.