Cover plants can be introduced in cropping systems to provide agroecosystem services, including weed control via competition for resources. There is currently no consensus on how to identify the best cover plant species, while trait-based approaches are promising for screening plant species due to their agroecosystem service provision potential. This study was carried out to characterize soil exploitation strategies of cover plant species in banana agroecosystems using a trait-based approach, and in turn identify cover plant species with a high weed control potential via competition for soil resources in banana cropping systems. A field experiment was conducted on 17 cover plant species, two weed species and two banana cultivars grown individually. Four functional traits were measured. Two of them (i.e., the size of the zone explored by roots and the root impact density) were used to characterize root system soil exploration patterns. Two other traits (i.e., specific root length and root diameter) were used to characterize resource acquisition within the soil zone explored by the roots. All studied traits exhibited marked variations among species. The findings suggested a trade-off between the abilities of species to develop a limited number of large diameter roots exploring a large soil zone versus many thin roots exploring a smaller soil zone. Three soil-resource exploitation strategies were identified among species: (i) with large diameter roots that explore a large soil zone; (ii) with small diameter roots and a high specific length that explore a smaller soil zone; and (iii) with a high total root-impact density and an intermediate specific root length that explore the uppermost soil layers. Interestingly, in our panel of species, no correlations with regard to belowground and aboveground strategies were noted: species with an acquisitive belowground strategy could display an acquisitive or a conservative aboveground strategy. The findings of this study illustr