A recent study suggests that the quasar HE 0450-2958 is hosted by a galaxy substantially fainter than that inferred from the correlation between black hole mass and bulge luminosity. As this result has significant bearings on galaxy and black hole evolution, we revisit the issue by performing an independent analysis of the data, using a two-dimensional image fitting technique. We indeed find no evidence of a host galaxy either, but due to the brightness of the quasar and uncertainties in the point-spread function, the limits are fairly weak. To derive an upper limit on the host galaxy luminosity, we perform simulations to deblend the quasar from the host under conditions similar to those actually observed. We find that the host galaxy has an absolute magnitude upper limit of -20<~MV<~-21, in good agreement with the previous determination. Since this limit is consistent with the value predicted from the current best estimate of the black hole mass, there is no compelling evidence that the quasar HE 0450-2958 has an abnormally underluminous host galaxy. We also show that, contrary to previous claims, the companion galaxy to HE 0450-2958 should not be regarded as an ultraluminous infrared galaxy.