An experiment was carried out at the High Rainfall Station of the International Institute of Tropical Agriculture (IITA) at Onne (4°43'N, 7°01'E, 10 masl), in southeastern Nigeria. The objectives of the study were (a) to assess the combining ability of Musa genotypes under alley and monocropping systems, (b) to estimate heritability of some quantitative traits of Musa species, and (c) to determine selection efficiency in each cropping system. The experimental population included about 560 progenies derived from crosses between tetraploid and diploid Musa hybrids. These were field established under alley and monocropping systems and evaluated for two consecutive cycles. Results indicated significant main effects of cropping system for days to flowering and plant height. The two-way interactions between cropping system x clone and crop cycle x clone was not significant. Ratio of GCA to SCA indicated the preponderance of additive gene effects in determining the inheritance of all traits under alley cropping, and days to flowering and bunch weight in the monocropping system. In a combined cropping system analysis, all the traits studied were highly heritable (H2 > 0.85). Furthermore, selection based on bunch weight of 10 kg or above identified 25 individuals under alley cropping, in contrast to 8 individuals under mono cropping. Three individuals were selected in both systems, giving an efficiency of 2.2 % and 35.6 % under alley and monocropping systems, respectively. Therefore, selection under the monocropping system would likely produce genotypes suitable for cultivation under alley cropping than the reverse option.