A systematic study of line shapes obtained by electrolyte electroreflectance (EER) and by automatic spectroscopic ellipsometry (ASE) on samples of mercury cadmium telluride (MCT) with varying densities of defects has been carried out. The results show that the two techniques yield in all cases the same value for the interband transition energies Ej. Because the values of the Ej are quantitatively related to the alloy composition x, it is clear that both ASE and EER can be used to carry out studies of the alloy composition. On the other hand, the values of the linewidths Γj yielded by the two techniques often do not agree. In those cases in which the etch pit density is less than 106/cm2, the values agree approximately; otherwise, EER yields much larger values. The difference between the two techniques is, of course, the use of an electric field modulated at very low frequencies (EER). We show that the modulating electric field causes an electrostriction and polarization of the defects in MCT. These effects, which have not been taken into account in previous theories of electroreflectance, are shown to give rise to first- and second-derivative line shapes proportional to the Stark shifts ΔEj and to a shift Δσ 2 related to the overall defect scattering strength, respectively. These new line shapes dominate the usual third-derivative line shape for defective materials. A simple interpretation of our experimental results suggests that the relative Stark shifts ΔEj are associated with plasticity, long-range strains and extended defects, and that the shift Δσ2 arises from polarizable charged defects. Thus, the treatment present here, by allowing one to determine values for ΔEj and Δσ2 as well as values for Γj in agreement with ASE, give important physical information not previously available on defects in MCT. The quantitative determination of the defect-related parameters has allowed us to establish a hierarchy of quality among all the epitaxial techniques by comparison to bulk single crystal samples.