The aim of this study was to estimate genetic parameters for beef fatty acids profile, carcass and meat traits in Nellore cattle. A total of 1616 Nellore bulls finished in feedlot (90 d), and slaughtered with an average of 24 mo of age were used. Samples from the Longissimus thoracis muscle were collected 48 h after slaughter. The proportion of saturated (SFA), monounsaturated (MUFA) and polyunsaturated fatty acids (PUFA), n-3 and n-6 were calculated. The total lipid concentration (LIP) was quantified using the Bligh and Dyer (1959) method. Fatty acids were quantified by gas chromatography (GC-2010 Plus - Shimadzu AOC 20i autosampler) using SP-2560 capillary column. The backfat thickness (BF) was measured using a caliper. Shear force (SF) was measured by the method of Wheeler et al. (1995). The (co)variances and genetic parameters were estimated using the restricted maximum likelihood method, considering an animal model (ssGBLUP), applying the remlf90 software. The animals were genotyped with the BovineSNP BeadChip (High-Density Bovine BeadChip) that contained 777,962 SNP markers. After quality control for genotypes, a total of 470,007 SNPs from 1,556 samples remained. The heritability estimates were low for SFA (0.09), MUFA 0.005, and moderate for PUFA (0.20), n3 (0.08) and n6 (0.20). For SFA, the genetic correlation estimates with SF (0.04) with BF (0.30), and LIP (0.27) were low. The genetic correlation estimates between MUFA with carcass and meat traits were low to high, being −0.05 with SF, −0.64 with BF, and 0.88 with LIP. The genetic correlation estimates between PUFA were low with SF (0.22), moderate with BF (−0.41) and LIP (−0.64). For n3, the genetic correlation estimates with were low with BF (−0.17) moderate with SF (−0.37) and with LIP (−0.40). The genetic correlation estimates between omega 6 with BF and SF were low (0.05 and 0.10, respectively), and high with LIP (−0.86). It is expected that the selection for carcass and meat traits would change the beef fatty acid composition.