Identification of transfer RNAs in animal mitochondrial genomes is important for many areas of genome analysis including phylogenetic reconstruction, understanding inheritance of disease, and identifying forensic materials. Animal mitochondrial tRNAs differ from the canonical tRNAs in both their secondary structure and level of conservation of nucleotide sequence and therefore, conventional tRNA or general RNA searching software cannot be used for identification and custom methods are required. Here we present the results of an experimental analysis of four different methods tested on a large dataset consisting of 5,720 tRNAs extracted from the entire set of complete animal mitochondrial genomes in GenBank. Methods were evaluated based on number of false negatives and false positives. Additionally, we present a new scoring scheme customized for animal mitochondrial tRNAs.