Abstract Background β‐ketothiolase (T2, gene symbol ACAT 1 ) deficiency is an autosomal recessive disorder, affecting isoleucine and ketone body metabolism. We encountered a patient ( GK 03) with T2 deficiency whose T2 mRNA level was <10% of the control, but in whom a previous routine cDNA analysis had failed to find any mutations. Genomic PCR ‐direct sequencing showed homozygosity for c.941‐9T>A in the polypyrimidine stretch at the splice acceptor site of intron 9 of ACAT 1 . Initially, we regarded this variant as not being disease‐causing by a method of predicting the effect of splicing using in silico tools. However, based on other findings of exon 10 splicing, we eventually hypothesized that this mutation causes exon 10 skipping. Methods cDNA analysis was performed using GK 03's fibroblasts treated with/without cycloheximide ( CHX ), since exon 10 skipping caused a frameshift and nonsense‐mediated mRNA decay ( NMD ). Minigene splicing experiment was done to confirm aberrant splicing. Results cDNA analysis using fibroblasts cultured with cycloheximide indeed showed the occurrence of exon 10 skipping. A minigene splicing experiment clearly showed that the c.941‐9T>A mutant resulted in transcripts with exon 10 skipping. There are few reports describing that single‐nucleotide substitutions in polypyrimidine stretches of splice acceptor sites cause aberrant splicing. Conclusion We showed that c.941‐9T>A induces aberrant splicing in the ACAT 1 gene. Our ability to predict the effects of mutations on splicing using in silico tools is still limited. cDNA analysis and minigene splicing experiments remain useful alternatives to reveal splice defects.