This theoretical and experimental study focuses on instability in magnetohydrodynamic liquid metal fusion blanket duct flows with inflectional velocity profiles, the understanding of which is crucial to predictions of transitions between flow regimes in blanket systems. Flow regimes strongly affect heat and material advection across a poloidal blanket duct, and in the hopes of enhancing our capability to predict flow regimes and transport behavior in a wide range of blanket conditions, this study strives to uncover the mechanism behind MHD duct flow destabilization. This research builds on a large volume of theoretical, numerical and experimental work suggesting bulk instability and a two-stage mechanism as the most likely culprit for flow destabilization in these systems and supplies a much needed set of detailed experimental data to help validate numerical simulations and improve predictions of critical flow parameters, which have historically varied across orders of magnitude depending on the assumptions used to determine stability criteria for these flows. The derivation of an exact analytical solution for a fully-developed laminar electrically driven MHD duct flow with two wall jets is presented, which provides useful velocity field data that can help identify the laminar regime and helps determine the conditions required for quasi-two-dimensional (Q2D) flow. This is followed by a description of quasi-two-dimensional simulations of a duct-like electrically driven MHD cavity flow performed in parallel with an experiment with the same geometry and boundary conditions. Results from these two efforts are compared with the analytical solution and one another as a validation of their data, and key details of the flow dynamics and flow statistics are explored and used to identify the different types of instability present under a variety of conditions. The combination of these three avenues of research are leveraged to determine exactly how instabilities form in this system and through what stages they evolve as the flow approaches a transition to Q2D turbulence. The results are then used to develop tools that can more easily detect these regimes of instability from simple duct flow diagnostics.