Using field-aligned ion density profiles from a physical-numerical model of the plasmasphere, we investigated the reliability of techniques which use field line resonance measurements to remotely sense the plasmaspheric mass density ρ, over L = 1.6--3.4. We find that the common assumption of some functional law for the variation of ρ along different field lines may lead to distorted profiles of the equatorial density ρo as inferred from the observed fundamental field line eigenfrequency. The tests show that for midlatitude field lines (2.3 < L < 3.4) a radial power law with an index m $cong$ 1 might be appropriate for a large variety of solar and geomagnetic conditions, and this index well represents the mass density dependence on the outer part of the field line. Indeed, an accurate integral expression of the field line eigenperiod demonstrates that the low-altitude plasma tube provides a negligible contribution to the eigenperiod. However, at lower latitudes, higher m values are necessary to obtain correct estimates of ρo. In this case m represents only an average index, not related to any particular variation of ρ along the field line. This effective index is also significantly dependent on solar activity conditions and local time. It turns out that inferred temporal variations of ρo at a given low L-shell can be significantly overestimated when a fixed functional dependence for the field-aligned mass density is assumed. The performance of more sophisticated techniques which use higher harmonics is also examined.