Genomic predictors offer an opportunity to overcome the limitations of classical selection against boar taint, and this study evaluated different approaches to obtain such predictors. Samples from 941 pigs were included in a design which was dominated by 421 sib pairs, each pair having an animal with a high and a low skatole concentration (>0.3 ¿ug/g). All samples were measured for skatole and androstenone and genotyped using the lllumina SNP 60K porcine lllumina beadchip. The accuracy of predicting phenotypes was assessed by cross-validation using six different genomic evaluation methods, GBLUP and five Bayesian methods. The range of accuracies obtained by different prediction methods was narrow for androstenone, between 0.29 (Bayes Lasso) and 0.31 (Bayes B), and wide for skatole, between 0.21 (GBLUP) and 0.26 (Bayes SSVS). Relative accuracies corrected for h2, were 0.54-0.56 and 0.750.94 for androstenone and skatole, respectively. The whole genome evaluation methods gave greater accuracy than using QTL alone (one SNP for androstenone and one SNP for skatole). Also, the dominance genetic variation was ignored in national evaluations, so we estimated the dominance genetic variance for androstenone and skatole using SNP information. For androstenone, GBLUP with dominance effects included captured substantial ratio of the dominance genetic variances (13%) in total variation. For skatole, more dominance genetic variance was captured by regional chromosomal heritability approach, particularly on chromosome 9, where the proportion of chromosomal dominance genetic variance in total dominance variance was 96%. The results demonstrate that GBLUP for androstenone is the simplest genomic technology to implement and one of the most accurate methods while more specialised models may be preferable for skatole. Dominance genetic effects included could provide additional source of genetic variation for both traits, therefore, it is worthwhile considering in genomic evaluations.