Mitigating the risk due to requirement volatility requires that we treat this as a systemic problem and understand what handles project managers can manipulate for management of project risk. In this paper we propose a systems view of this problem using existing theory on user involvement and qualitative research conducted among experienced project managers. Augmenting an existing model of software project dynamics with elements of our understanding, we show that requirement volatility can occur as an outcome of uncontrolled user involvement defeating the very purpose for which user involvement is solicited. Project managers normally treat user involvement as a pre-requisite for obtaining software project success. What-if analysis with the model generates patterns of requirement volatility reported in literature. These insights into the relationship among user involvement, requirement volatility, and project performance are expected to assist project managers in devising user management strategies for controlling project risk.