Magnetorheological (MR) dampers can experience large temperature changes as a result of heating caused by energy dissipation, but control systems are often designed without consideration of this fact. Furthermore, due to the highly nonlinear behavior of MR dampers, many control strategies have been proposed and it is difficult to determine which is the most effective. This paper aims to address these issues through a numerical and experimental study of an MR mass isolator subject to temperature variation. A dynamic temperature dependant model of an MR damper is first developed and validated. Control system experiments are then performed using hardware-in-the-loopsimulations. Proportional, PID, gain scheduling, and on/ off control strategies are found to be equally affected by temperature variation. Using simulations incorporating the temperature dependant MR damper model, it is shown that this is largely due to a change in fluid viscosity and the associated movement of the lower `clipped optimal' control bound. This zero-volts condition determines how close any controller can perform to the ideal semiactive case, thus all types of controller are affected. In terms of relative performance, proportional and PID controllers perform equally well and outperform the on/off and gain scheduling strategies. Gain scheduling methods are superior to on/off control.