PartitionFinder is a new tool for selecting optimal partitioning schemes and substitution models in phylogenetic analyses. It allows users to compare millions of partitioning schemes efficiently using information-theoretic metrics like AIC, BIC, and AICc. The tool is implemented in Python and is freely available. PartitionFinder improves upon previous methods by objectively selecting partitioning schemes, which can significantly enhance the accuracy of phylogenetic reconstructions. The algorithm uses a greedy heuristic approach to search for optimal partitioning schemes, reducing computational demands compared to exhaustive searches. It also allows for the selection of linked or unlinked branch lengths between subsets. PartitionFinder has been tested on various data sets and has shown superior performance compared to ad hoc and hierarchical clustering methods. The results indicate that a priori partitioning schemes are often suboptimal, while PartitionFinder consistently finds better partitioning schemes. The tool is particularly useful for large data sets where exhaustive searches are not feasible. PartitionFinder is an open-source program that facilitates the objective selection of partitioning schemes, leading to improved phylogenetic analyses.PartitionFinder is a new tool for selecting optimal partitioning schemes and substitution models in phylogenetic analyses. It allows users to compare millions of partitioning schemes efficiently using information-theoretic metrics like AIC, BIC, and AICc. The tool is implemented in Python and is freely available. PartitionFinder improves upon previous methods by objectively selecting partitioning schemes, which can significantly enhance the accuracy of phylogenetic reconstructions. The algorithm uses a greedy heuristic approach to search for optimal partitioning schemes, reducing computational demands compared to exhaustive searches. It also allows for the selection of linked or unlinked branch lengths between subsets. PartitionFinder has been tested on various data sets and has shown superior performance compared to ad hoc and hierarchical clustering methods. The results indicate that a priori partitioning schemes are often suboptimal, while PartitionFinder consistently finds better partitioning schemes. The tool is particularly useful for large data sets where exhaustive searches are not feasible. PartitionFinder is an open-source program that facilitates the objective selection of partitioning schemes, leading to improved phylogenetic analyses.