Introduction. Bananas are traditionally propagated by suckers: nevertheless, the advantage of using tissue-cultured plants for banana production has been demonstrated in recent years. Despite the large-scale use of micropropagated plants in the subtropical regions, little information is available on the performance of these crops in tropical countries. So we compared the yield potential of some Cavendish banana cultivars propagated either by traditional suckers or by tissue culture. Materials and methods. A randomized complete block experiment was established in a commercial plantation in Cameroon to compare the performance of three tissue-cultured banana plants of Grand Nain, Williams and Zelig cultivars and two conventionally-propagated plants of Grand Nain and Williams. Morphological, phenological and yield component data were collected for six crop cycles. Results. All the plants grew vigorously and produced quality exportable fruits. However, the tissue-cultured plants consistently produced heavier bunches and higher yields than the conventionally-propagated plants. The highest yields (65.92 t·ha–1 for Grand Nain, 64.60 t·ha–1 for Williams and 65.58 t·ha–1 for Zelig) were obtained for the tissue-cultured plants in the third crop cycle compared with 51.43 t·ha–1 and 59.02 t·ha–1 obtained for the traditionally-propagated Grand Nain and Williams plants, respectively. There were no significant yield differences between the plants of different origin in the second and fourth cycles, respectively. However, the tissue-cultured Williams and Zelig plants outperformed the other plants in the sixth crop cycle. Conclusion. The tissue-cultured plants of the three Cavendish cultivars studied performed better than the plants resulting from conventional suckers. The best cultivar for commercial plantation used, that is located in a tropical environment, was Williams. In general, to obtain an optimum benefit, we would recommend the use of tissue-cultured plants over a period of 6–7 years with good management practices.