The release of new, sweet-acid banana varieties resistant to Sigatoka and Fusarium wilt is important for domestic markets in tropical and subtropical countries. Common current breeding strategies consist of selecting tetraploid AAAB new hybrids directly from crosses between AAB sweet-acid varieties pollinated with AA clones carrying resistance genes. However, this cross pathway is hampered by low gamete fertility and the rare occurrence of desired 2N gametes on the AAB female side (N= X = 33 chromosomes). We propose an alternative pathway which aims to create new triploid hybrids directly from AB landraces. Natural AB clones are sterile but their AABB tetraploid counterparts obtained by colchicin treatment are fertile. This gamete fertility was made profitable in crosses with AA and BB accessions to generate AÁB and ABB hybrids. We present here our results on the agronomic value of various progenies involving an in vitro synthesized tetraploid Kunnan (AABB) and several AA and ^^ clones. These first results suggest a very high potential of this new strategy for the release of well-performing new hybrids combining productivity, disease resistance and fruit quality. Hybrids with a high added-value, issued from this strategy could be made available for evaluation within the ProMusa network.