The OncoScan CNV Plus (OC+) assay is a single-nucleotide polymorphism-based microarray designed to detect genome-wide copy-number imbalance and loss of heterozygosity. This assay is also designed to detect 64 mutations in 9 cancer genes (BRAF, EGFR, IDH1, IDH2, KRAS, NRAS, PIK3CA, PTEN and TP53). We evaluated the accuracy of OC+ mutation calling in comparison to mutation testing by a 118-gene targeted amplicon-based neuro-oncology next-generation sequencing (NGS) panel for three brain tumor clinically relevant mutations: IDH1 R132H, IDH2 R172K and BRAF V600E. Cases included 636 brain tumor samples with tumor purity of at least 20%. IDH1 R132H mutation was called by OC+ in 138 cases; 137 cases were confirmed by NGS (sensitivity: 95.8%) and the single false positive case had an IDH1 R132L mutation. Among the 498 cases called negative for IDH1 R132H mutation by OC+, 6 were false negatives and shown to harbor an IDH1 R132H mutation by NGS (specificity: 99.8%). False negative cases had low variant allele frequency (20% or less), low DNA input and/or suboptimal OC+ quality metrics. All IDH2 R172K (n=14) and BRAF V600E (n=20) mutations called by OC+ were confirmed by NGS; no additional cases were detected by NGS, with sensitivity and specificity of 100% for calling of these 2 mutations by OC+. Our results indicate that OC+ mutation calling is accurate for IDH2 R172K and BRAF V600E mutations, and very specific but relatively less sensitive for IDH1 R132H mutation, suggesting that OC+ may be suitable for direct mutation calling for these three mutations. The OncoScan CNV Plus (OC+) assay is a single-nucleotide polymorphism-based microarray designed to detect genome-wide copy-number imbalance and loss of heterozygosity. This assay is also designed to detect 64 mutations in 9 cancer genes (BRAF, EGFR, IDH1, IDH2, KRAS, NRAS, PIK3CA, PTEN and TP53). We evaluated the accuracy of OC+ mutation calling in comparison to mutation testing by a 118-gene targeted amplicon-based neuro-oncology next-generation sequencing (NGS) panel for three brain tumor clinically relevant mutations: IDH1 R132H, IDH2 R172K and BRAF V600E. Cases included 636 brain tumor samples with tumor purity of at least 20%. IDH1 R132H mutation was called by OC+ in 138 cases; 137 cases were confirmed by NGS (sensitivity: 95.8%) and the single false positive case had an IDH1 R132L mutation. Among the 498 cases called negative for IDH1 R132H mutation by OC+, 6 were false negatives and shown to harbor an IDH1 R132H mutation by NGS (specificity: 99.8%). False negative cases had low variant allele frequency (20% or less), low DNA input and/or suboptimal OC+ quality metrics. All IDH2 R172K (n=14) and BRAF V600E (n=20) mutations called by OC+ were confirmed by NGS; no additional cases were detected by NGS, with sensitivity and specificity of 100% for calling of these 2 mutations by OC+. Our results indicate that OC+ mutation calling is accurate for IDH2 R172K and BRAF V600E mutations, and very specific but relatively less sensitive for IDH1 R132H mutation, suggesting that OC+ may be suitable for direct mutation calling for these three mutations.