This　paper　addresses　the　issues　related　to　the　design　of　robust　controller　using　genetic　algorithms　(GA)　for　lightweight,　one-link　flexible　manipulators　working　under　dynamic　environments　and　other　uncertain　influences.　By　selecting　sensitivity　weight　functions　properly　using　the　GA　method,　a　mixed　sensitivity　H　infinity　controller　is　developed　to　ensure　robustness　of　manipulator　control　systems　for　varying　payloads　and　other　modeling　uncertainties.　Numeric　simulation　has　been　conducted　and　the　results　have　demonstrated　the　effectiveness　of　the　proposed　method.