Canada has two core payment systems for processing funds transfers between financial institutions: the Large Value Transfer System (LVTS) and the Automated Clearing Settlement System (ACSS). These will be replaced in the next few years by three new systems: Lynx, the Settlement Optimization Engine (SOE) and the Real-Time Rail (RTR). We employ historical LVTS and ACSS data and use the discrete choice demand estimation approach to uncover end users’ and financial institutions’ preferences when deciding which payment instruments and payment systems, respectively, to use. Based on the estimated preferences revealed, we conduct various counterfactual analyses to predict the volume and value shares of the future payment systems. The results show that small-value LVTS payments will likely migrate to the SOE. Also, in the short run, about CA$10 000 billion of LVTS and ACSS payments per year is anticipated to migrate to the RTR if not subject to maximum transaction values. These migration patterns raise important policy questions, such as whether the future systems should be subject to value caps and/or higher collateral requirements.