Meeting future demands of food security will require enhanced rice production that is more environmentally sustainable. To achieve this it is important to know the genetic and molecular mechanism controlling the root traits. High throughput phenotyping which can keep pace with genotyping is needed, but for many researchers this needs to be cheap as well as meaningful. Here a very simple, low cost and reliable method of assessing root depth of seedling using a layer of diuron-soaked filter paper buried 25 cm deep in a soil-filled box has been developed which is suitable for screening of hundreds of accessions. The assumption is that deep-rooting plants die quicker. This method was then used to screen five established rice panels. Deep rooted cultivars were screened from a panel of an aus panel from IRRI and a panel of Brazilian and Japanese cultivars by using this method. Root QTLs were detected by using bi-parental mapping population and GWA study was performed in two panels, the rice diversity panel (RDP-1) and Bengal Assam Association Population. Assessing 139 RILs from Bala x Azucena bi-parental population revealed heritability of 55% for herbicide symptoms where eleven QTLs were detected, many of which were co-localised with previously reported root QTLs in this population. A GWA study was performed using RDP1) of 356 accessions with 44k SNP markers. Analysis revealed 17% of phenotypic variation of herbicide score was attributable to rice sub-population where the aus showed the deepest rooting systems. A number of QTLs have been identified and a number of positional candidate gene lists were produced. A further 298 cultivars from Bengal and Assam were screened and GWA was performed using 2 M SNP database available from sequencing. ANOVA revealed 37% variation for herbicide score explained by genotype. Soil-filled rhizotron were used to assess 12 of these cultivars, revealing strong xx correlations between deep root traits and herbicide score, confirming the reliability of this method. GWA revealed a number of significant SNPs associated with the traits in this population. Finally a set of mutant gene (LOC_Os09g31478, LOC_Os05g40330, LOC_Os11g34140) which are functional candidate gene for root growth QTLs were studied. Here hydroponic phenotypic screening approach were used to identify the T-DNA mutant lines. However, no convincing mutants were revealed. The herbicide screening method has been shown to be a quick and robust system for the assessment of deep rooting rice plants in soil. This method can now be used for screening large number of cultivars and the identification of QTLs and candidate genes.