HHblits is an open-source, iterative protein sequence search tool that uses profile hidden Markov models (HMMs) to represent query and database sequences. It is designed to be faster and more sensitive than PSI-BLAST, with 50–100% higher sensitivity and more accurate alignments. HHblits employs a discretized-profile prefilter to reduce the number of full HMM-HMM alignments, making it significantly faster while maintaining high sensitivity. The tool can be used to build multiple sequence alignments (MSAs) for downstream analysis and prediction methods, such as de novo protein structure prediction. HHblits is particularly useful for searching large databases like UniProt and NCBI's nonredundant (nr) database, and it has been shown to improve the accuracy of secondary structure predictions and fold predictions for Pfam families. The tool is available as source code and executable packages for various platforms and can be accessed through the HHblits server.HHblits is an open-source, iterative protein sequence search tool that uses profile hidden Markov models (HMMs) to represent query and database sequences. It is designed to be faster and more sensitive than PSI-BLAST, with 50–100% higher sensitivity and more accurate alignments. HHblits employs a discretized-profile prefilter to reduce the number of full HMM-HMM alignments, making it significantly faster while maintaining high sensitivity. The tool can be used to build multiple sequence alignments (MSAs) for downstream analysis and prediction methods, such as de novo protein structure prediction. HHblits is particularly useful for searching large databases like UniProt and NCBI's nonredundant (nr) database, and it has been shown to improve the accuracy of secondary structure predictions and fold predictions for Pfam families. The tool is available as source code and executable packages for various platforms and can be accessed through the HHblits server.