Roots are key organs for plants as transporters of water and nutrients. However, they are complicated to observe. To develop rice varieties with root systems adapted to the target environments (e.g. with a high number of deep, thick and highly branched roots for the upland ecosystem) without direct observation, a better knowledge of the genes involved is necessary. So far, only three genes influencing root architecture or nutrient uptake have been cloned in rice. Whole genome association mapping (WGAM) offers a fast access to new quantitative trait loci (QTLs) with a better resolution than conventional QTL mapping. Results on several panels are presently available. However, WGAM relies on statistics and its results need further validation using evidences from other approaches, notably functional genomics. The effects of genes with function on root development validated in Arabidopsis are being studied in rice (e.g. SHR and SCR or the family of leucine-rich repeat receptor-like kinases in our lab). We will present a summary of all these results.