The structure of strongly interacting, charged dispersions is examined using the Ornstein-Zernike integral-equation formalism with thermodynamic consistency conditions (i.e., with the so-called HMSA closure conditions). It is shown that for highly charged dispersions interacting through screened Coulombic (Yukawa) interactions this integral-equation approach predicts the static structure factor S(q) in excellent agreement with Monte Carlo results and that it is better than the rescaled mean-spherical approximation. The Sogami potential (which predicts Coulombic attraction under identical physicochemical conditions) is also considered here as a model potential for representing soft, long-ranged interactions with an attractive component. None of the integral-equation formalisms, including the HMSA theory and the reference hypernetted-chain (RHNC) theory, leads to a sufficient level of thermodynamic consistency for the latter potential, and the results deviate noticeably from the Monte Carlo results. We further demonstrate that fitting the experimentally observed structure factors in the neighborhood and beyond the primary peak in S(q) could lead to inaccurate conclusions concerning the nature of interparticle forces, particularly in the case of soft, long-range interactions.