Over the past 60 years the internal rate of return (IRR) has become a major tool in investment evaluation. Many executives prefer it to net present value (NPV), presumably because they can more easily comprehend a percentage measure. This article demonstrates that, except in the rare case of an investment that is followed by a single cash return, IRR suffers from a definitional quandary. Is it an intrinsic measure, defined only in terms of itself, or is it defined by the efforts of active investors? Additionally, the article explains significant problems with the measure--reinvestment issues, multiple IRRs, timing problems, problems of choice among unequal investment opportunities, and practical difficulties with multiple discount rates. IRR is a blind guide because its definition is in doubt and because of its many practical problems.