It has recently been proposed, on the basis of eclipse-timing data, that the eclipsing polar cataclysmic variable HU Aquarii is host to at least two giant planets. However, that result has been called into question based upon the dynamical stability of the proposed planets. In this work, we present a detailed re-analysis of all eclipse-timing data available for the HU Aquarii system, making use of standard techniques used to fit orbits to radial-velocity data. We find that the eclipse timings can be used to obtain a two-planet solution that does not require the presence of additional bodies within the system. We then perform a highly detailed dynamical analysis of the proposed planetary system. We show that the improved orbital parameters we have derived correspond to planets that are dynamically unstable on unfeasibly short time-scales (of the order of 104 yr or less). Given these results, we discuss briefly how the observed signal might in fact be the result of the intrinsic properties of the eclipsing polar, rather than being evidence of dynamically improbable planets. Taken in concert, our results highlight the need for caution in interpreting such timing variations as being planetary in nature.