This talk discusses the use of an end-user computing environment to engage students with programming practice and computational thinking using the context of musical creativity. The talk will detail (with live demonstrations) interactions and lessons using the Manhattan environment, which combines music editing with a scalable level of programming – from simple formulas or code fragments embedded in musical patterns to more advanced algorithms and generative music. The environment enables a wide range of fundamental and advanced computing concepts to be taught visually and musically. Discussions will focus on the importance of supporting intrinsic motivation, scalable challenge, and personal creativity, in tackling complex learning domains like computer science and music theory, and are supported by data and student feedback from classroom use. Recent and future developments, exploring realtime collaboration and connectivity will also be discussed in the context of supporting social learning and extrinsic motivation, and broadening use of the tool.