Dysregulation of fatty acids metabolism has been associated with the risk of osteoarthritis (OA), yet current evidence from epidemiological or genetic studies remains inconclusive. We aimed to investigate the phenotypic association and genetic architecture between total fatty acids (TotFA), saturated fatty acids (SFA), monounsaturated fatty acids (MUFA), polyunsaturated fatty acids (PUFA), and OA. Leveraging individual-level data from the UK Biobank, combined with the hitherto largest genome-wide association studies of fatty acids (N = 136 016) and OA (N = 826 690) in European individuals, we implemented a comprehensive analytical framework. This included observational and genetic analyses, incorporating phenotypic associations, genetic correlations, cross-trait meta-analysis, enrichment analysis and Mendelian randomization (MR). Observational analysis identified SFA as a risk factor, while MUFA and PUFA as protective factors for OA. Despite a lack of genome-wide genetic correlation, statistically significant local signals were detected within three specific genomic regions. Cross-trait meta-analysis identified 68 pleiotropic loci shared between fatty acids and OA, of which nine were novel. Enrichment analysis revealed the shared genes were enriched in lipoprotein metabolism, immune response, and inflammation regulation pathways. Two-sample MR provided evidence for a causal relationship of MUFA and PUFA on OA that survived false discovery rate correction. This study supports associations between circulating fatty acids and OA, with MUFA and PUFA exerting a protective role. Our findings provide new perspectives into OA prevention especially regarding the potential dietary interventions.