TrimAl is a tool for automated alignment trimming in large-scale phylogenetic analyses. It is designed to improve the quality of multiple sequence alignments (MSA) by removing poorly aligned regions. TrimAl considers various parameters, such as the proportion of sequences with gaps, amino acid similarity, and consistency across different alignments, to select the most reliable positions in the alignment. It can automatically adjust parameters to optimize the signal-to-noise ratio, making it suitable for large-scale phylogenomic analyses. TrimAl is written in C++ and is available for download and online use. It supports various alignment formats and can process large datasets efficiently. The tool uses different scoring methods, including gap scores, similarity scores, and consistency scores, to evaluate the quality of each column in the alignment. Based on these scores, TrimAl can trim alignments to retain only the most conserved regions. TrimAl has been tested on various datasets and has shown improved phylogenetic accuracy compared to other tools like Gblocks. When used with automated parameter selection, TrimAl produced Maximum Likelihood trees of equal or better quality than those from complete alignments. It also outperformed Gblocks in most scenarios, especially when using default parameters. TrimAl was tested on real datasets, including the Human Phylome project, demonstrating its effectiveness on large-scale phylogenetic data. It has been used in pipelines to reconstruct gene trees and is available through PhylomeDB. The tool is supported by a wiki page and online documentation, providing detailed information on its usage and performance. Overall, TrimAl is a powerful tool for improving the accuracy of phylogenetic analyses by efficiently trimming alignments.TrimAl is a tool for automated alignment trimming in large-scale phylogenetic analyses. It is designed to improve the quality of multiple sequence alignments (MSA) by removing poorly aligned regions. TrimAl considers various parameters, such as the proportion of sequences with gaps, amino acid similarity, and consistency across different alignments, to select the most reliable positions in the alignment. It can automatically adjust parameters to optimize the signal-to-noise ratio, making it suitable for large-scale phylogenomic analyses. TrimAl is written in C++ and is available for download and online use. It supports various alignment formats and can process large datasets efficiently. The tool uses different scoring methods, including gap scores, similarity scores, and consistency scores, to evaluate the quality of each column in the alignment. Based on these scores, TrimAl can trim alignments to retain only the most conserved regions. TrimAl has been tested on various datasets and has shown improved phylogenetic accuracy compared to other tools like Gblocks. When used with automated parameter selection, TrimAl produced Maximum Likelihood trees of equal or better quality than those from complete alignments. It also outperformed Gblocks in most scenarios, especially when using default parameters. TrimAl was tested on real datasets, including the Human Phylome project, demonstrating its effectiveness on large-scale phylogenetic data. It has been used in pipelines to reconstruct gene trees and is available through PhylomeDB. The tool is supported by a wiki page and online documentation, providing detailed information on its usage and performance. Overall, TrimAl is a powerful tool for improving the accuracy of phylogenetic analyses by efficiently trimming alignments.