INFORMATION PERSPECTIVE ON TURBULENCE Turbulence is the proverbial fruit y of complex systems. Information theory [1, 2, 3, 4], and the more recently developed computational mechanics , provide quantitative tools for quantifying \complexity" in terms of predictability. This thesis is concerned with using these tools with experimental turbulence data. It is this predictability that we are most interested in, and it will serve as a unifying theme for this thesis. The focus will be on two-dimensional turbulence [32, 33], but other (mostly turbulent) systems are treated as well. The role of information theory in physical systems in general will also be discussed. This framework is general and easily applicable to almost any other kind of system, as will be shown.