The article introduces ProtTest, a Java program designed to select the best-fit model of amino acid replacement for a given protein sequence alignment. The program is based on the Phylm program for maximum likelihood optimizations and uses the PAL library for handling protein alignments and trees. ProtTest supports various empirical models, including the WAG, mtREV, Dayhoff, JTT, VT, Blossem62, CpREV, and RrREV matrices, with options for +F, +G, and +I constraints. The program can be run using a GUI, command-line, or web interface, and it provides tools for model selection based on the Akaike Information Criterion (AIC), corrected AIC (AICc), and Bayesian Information Criterion (BIC). The article also discusses the importance of appropriate models in protein evolution studies and the methods for model selection, including the use of information criteria to balance accuracy and simplicity.The article introduces ProtTest, a Java program designed to select the best-fit model of amino acid replacement for a given protein sequence alignment. The program is based on the Phylm program for maximum likelihood optimizations and uses the PAL library for handling protein alignments and trees. ProtTest supports various empirical models, including the WAG, mtREV, Dayhoff, JTT, VT, Blossem62, CpREV, and RrREV matrices, with options for +F, +G, and +I constraints. The program can be run using a GUI, command-line, or web interface, and it provides tools for model selection based on the Akaike Information Criterion (AIC), corrected AIC (AICc), and Bayesian Information Criterion (BIC). The article also discusses the importance of appropriate models in protein evolution studies and the methods for model selection, including the use of information criteria to balance accuracy and simplicity.