Arachis hypogaea, the cultivated groundnut is a tetraploid with an AABB genomic constitution. The available literature on the origin of groundnut reveals that there is general agreement that the cultivated groundnut has evolved from the wild tetraploid species A. monticola, with which it crosses freely to produce fertile hybrids. However, the issue of actual diploid ancestors of A. monticola is still unresolved. Both cytogenetic and molecular evidences support A. duranensis being the most probable progenitor and donor of the A genome to A. hypogaea. For the B genome, the cytogenetic evidence suggests A. batizocoi to be the most probable progenitor, but the RFLP banding pattern indicates that A. batizocoi is more distantly related to A. hypogaea than other species of section Arachis. RFLP banding pattern indicates A. ipaensis to be one of the closest species to A. hypogaea and the possible donor of the B genome. The present article critically analyzes the available data, which suggests that until an amphidiploid is produced synthetically between A. duranensis × A. ipaensis and crossed successfully with A. hypogaea to produce a fertile hybrid, this issue would remain unresolved. A. batizocoi would remain the most probable donor of the B genome because of its directly demonstrable cytogenetic affinity.