BACKGROUND: Fat content (FC) and fatty acids (FA) are the most important traits in shea tree breeding, controlled by several genes with relatively small effects. Therefore, determining the genes involved in the biosynthesis of such traits is crucial for improving oil quantity and quality and for the domestication process of the species. To identify the quantitative trait nucleotides (QTNs) controlling FC and FA, we conducted a multi-locus genome-wide association study (GWAS) using six multi-locus GWAS methods for FC and FA in 122 superior shea trees (SSTs). SSTs were genotyped using DArTseq, resulting in 7,559 non-redundant single nucleotide polymorphism markers. RESULTS: Fat content varied from 36 to 58% with a mean of 50%. Fatty acid composition was 51.26 ± 4.21, 38.76 ± 4.67, 6.45 ± 0.76 and 3.53 ± 0.52% for oleic, stearic, linoleic and palmitic acids, respectively. A very high negative correlation coefficient (-0.98) was found between stearic and oleic acids. A total of 47 significant QTNs associated with fat-related traits were detected by the GWAS methods. Among these QTNs, 25 were identified as common QTNs based on their detection by multiple GWAS methods. Using the superior allele information of the 4 common QTNs associated with fat content in 17 high-fat and 21 low-fat SSTs, we found a higher percentage of superior alleles in SSTs with high FC (47.1%) than in SSTs with low FC (14.3%). Pathway analysis of the common QTNs identified 24 potential candidate genes likely involved in the biosynthesis of FC and FA composition in shea tree seeds. CONCLUSIONS: These findings will contribute to the discovery of the polygenic networks controlling FC in shea tree, improve our understanding of the genetic basis and regulation of FC, and be useful for molecular breeding of high-fat shea tree cultivars.