The Pfam protein families database is a widely used resource for protein family and domain analysis. This article describes major updates in the latest release (version 24.0), including the adoption of HMMER3, which is significantly faster and more sensitive than the previous version. Pfam 24.0 contains 11,912 families, with many updated over the past two years. Pfam is available via servers in the UK, USA, and Sweden. Pfam is a comprehensive database of conserved protein families, used extensively in biological sciences. It is also used in structural biology for identifying new targets for structure determination. Pfam has been designed to scale with the growth of new protein sequences. It uses seed alignments to build profile hidden Markov models (HMMs) for sensitive and accurate sequence searches. The move to HMMER3 necessitated changes to Pfam, including the use of a new version of the profile HMM software. Pfam 24.0 has a 24% increase in the number of families compared to Pfam 23.0. The database has also increased coverage, with sequence coverage rising from 75.15% in Pfam 23.0 to 75.15% in Pfam 24.0. Residue coverage also increased, from 51.22% to 53.18%. Pfam 24.0 includes 2871 new Pfam-A families and 236 deleted families. The database has also improved sensitivity, with HMMER3 providing more accurate results. Pfam 24.0 has also increased the number of distinct domain architectures, with a 67% increase from 48,634 to 72,629. Pfam 24.0 also includes 142,303 Pfam-B families, which are automatically generated families. The database has also improved its website features, including interactive sequence searches and new family page features. Pfam 24.0 also includes new features such as DAS ontology compliance and TreeFam links. The database has also improved its scientific annotations, with new features for retrieving information from BioLit and TOPSAN. The Pfam user community has also benefited from the new version, with the ability to run Pfam locally using the pfam_scan tool. The new version of pfam_scan runs significantly faster than the old version, with improved features and formats. The database has also improved its communication with the Pfam community through the Xfam blog. Overall, the changes to Pfam have significantly improved the database's performance, sensitivity, and coverage, making it a more comprehensive and accurate resource for protein family analysis.The Pfam protein families database is a widely used resource for protein family and domain analysis. This article describes major updates in the latest release (version 24.0), including the adoption of HMMER3, which is significantly faster and more sensitive than the previous version. Pfam 24.0 contains 11,912 families, with many updated over the past two years. Pfam is available via servers in the UK, USA, and Sweden. Pfam is a comprehensive database of conserved protein families, used extensively in biological sciences. It is also used in structural biology for identifying new targets for structure determination. Pfam has been designed to scale with the growth of new protein sequences. It uses seed alignments to build profile hidden Markov models (HMMs) for sensitive and accurate sequence searches. The move to HMMER3 necessitated changes to Pfam, including the use of a new version of the profile HMM software. Pfam 24.0 has a 24% increase in the number of families compared to Pfam 23.0. The database has also increased coverage, with sequence coverage rising from 75.15% in Pfam 23.0 to 75.15% in Pfam 24.0. Residue coverage also increased, from 51.22% to 53.18%. Pfam 24.0 includes 2871 new Pfam-A families and 236 deleted families. The database has also improved sensitivity, with HMMER3 providing more accurate results. Pfam 24.0 has also increased the number of distinct domain architectures, with a 67% increase from 48,634 to 72,629. Pfam 24.0 also includes 142,303 Pfam-B families, which are automatically generated families. The database has also improved its website features, including interactive sequence searches and new family page features. Pfam 24.0 also includes new features such as DAS ontology compliance and TreeFam links. The database has also improved its scientific annotations, with new features for retrieving information from BioLit and TOPSAN. The Pfam user community has also benefited from the new version, with the ability to run Pfam locally using the pfam_scan tool. The new version of pfam_scan runs significantly faster than the old version, with improved features and formats. The database has also improved its communication with the Pfam community through the Xfam blog. Overall, the changes to Pfam have significantly improved the database's performance, sensitivity, and coverage, making it a more comprehensive and accurate resource for protein family analysis.