Imaging internal multiples properly has the potential to enhance the quality of the migrated images since they may illuminate subsurface zones that are poorly illuminated by single-scattering energy, such as vertical and nearly vertical faults, and salt flanks. The generalized interferometric multiple imaging procedure aims to migrate internally multiply-scattered energy. Imaging first-order internal multiples using this procedure is composed of a data extrapolation based datuming step, an interferometric crosscorrelation datuming step followed by a zero-lag cross-correlation imaging condition with the forward extrapolated datumed data. However, this procedure yields migrated images that suffers from low spatial resolution, and migration artifacts due to the datuming steps based on cross-correlation, cross-talk noise, and band-limited nature of the source wavelet. We propose a least-squares generalized interferometric multiple imaging procedure, in which we replace the first two datuming steps by least-squares datuming ones. The proposed framework would suppress the cross-talk noise and migration artifacts, and deconvolve the source wavelet. Therefore, it would enhance the resolution of the subsurface reflectivity distribution illuminated by first-order internal multiples. Application to two synthetic datasets improved the illumination of subsurface scatterers and the high-resolution delineation of a vertical fault plane using first-order internal multiples.