The notion that plants harbor diversified metabolites could serve as sources of human nutrition and as effectors of plant responding to environmental conditions has attracted increasing attention to explore the functional genes and the interconnected pathways underlying the content variation of these metabolites. Forward-genetics-metabolomics methods, like mGWAS and mQTL, have been proved effective in providing fruitful loci affecting metabolite contents, wherein candidate genes could be identified. However, the statistical power of mGWAS or mQTL is somehow constrained by constitutional flaws of respective mapping methodologies. Here, we present the combination of mGWAS and mQTL in wheat, wherein metabolic routes could be unveiled by either manner and further by combining both strategies. Specifically, a total of 480 and 293 loci were respectively generated from mGWAS and mQTL, in which 23 of them were considered to be co-mapped by both populations. Benefiting from the loci mapped from the two populations, we have revealed metabolic routes from either mGWAS or mQTL loci by enzymatically validating candidate genes that encode glycosyl-transferases and methyl-transferases in the flavonoid pathway, and a metabolic pathway was further disclosed by combining both approaches. Moreover, validation of two genes that encode a MYB transcription factor and a methyl-transferase by transgenic wheat further consolidated the effectiveness of the candidate gene identification. In view of the flavonoids were constantly considered to impart human health and contribute plant growth regulation, we propose our survey exemplified that the forward-genetics-metabolomics approaches including multiple populations could be more frequently applied for delineating metabolic pathways in common wheat, which will ultimately replenish molecular resources for wheat crop improvement.