Isostatic equilibrium is commonly defined as the state achieved when there are no lateral gradients in hydrostatic pressure, and thus no lateral flow, at depth within the lower viscosity mantle that underlies a planetary body's outer crust. In a constant-gravity Cartesian framework, this definition is equivalent to the requirement that columns of equal width contain equal masses. Here we show, however, that this equivalence breaks down when the spherical geometry of the problem is taken into account. Imposing the "equal masses" requirement in a spherical geometry, as is commonly done in the literature, leads to significant lateral pressure gradients along internal equipotential surfaces and thus corresponds to a state of disequilibrium. Compared with the "equal pressures" model we present here, the equal masses model always overestimates the compensation depth-by similar to 27% in the case of the lunar highlands and by nearly a factor of 2 in the case of Enceladus. Plain Language Summary "Isostasy" is the principle that, just as an iceberg floats on the water, crustal rocks effectively float on the underlying higher density mantle, which behaves essentially like a fluid on geologic timescales. Although there are subtle inconsistencies among the various ways isostasy can be defined, they have not been historically problematic for bodies like the Earth, where the crust is thin compared with the overall radius. When the thickness of the crust is a nonnegligible fraction of a planetary body's radius, however, it becomes important to take the spherical geometry into account. In this case, the inconsistencies in the definitions can lead to significant discrepancies. Here we argue that one of the most commonly used approaches, which requires equal width columns to contain equal masses, always results in overestimating the crustal thickness. In particular, we suggest that the lunar and Martian highlands crustal thickness may have been overestimated by similar to 27% and similar to 10%, respectively, and that the ice shell thickness for Saturn's small icy moon Enceladus may have been overestimated by nearly a factor of 2.