Our attempts to control risks associated with our software projects have given rise to a multitude of formal and informal classification schemes for factors that can negatively impact a software project. Despite this work, we have seen slow improvement in project success rates in an increasingly complicated domain with overlapping, inconsistently applied definitions of our major concepts. The research agenda detailed here promises to review past work regarding the conceptualization of software development project risk, analyze it using current best practices for taxonomy development and evaluation, and then re-organize it into an updated taxonomy intended to help researchers and practitioners better understand the domain and improve project success rates.