JASPAR is an open-access database containing high-quality, matrix-based transcription factor (TF) binding site profiles for multicellular eukaryotes. The profiles are derived from experimentally validated nucleotide sequences that bind TFs. The database includes a web interface for browsing, searching, and subset selection, an online sequence analysis tool, and programming tools for genome-wide and comparative genomic analysis. JASPAR is available at http://jaspar.cgb.ki.se. The database contains 111 profiles, derived from published collections of experimentally defined TF binding sites. Profiles were curated based on quality assessment criteria, including experimental validation, absence of arbitrary filtering, unambiguous sequence availability, and sufficient binding site sequences. Profiles were aligned using the ANNSpec program, which uses artificial neural networks and Gibbs sampling for pattern discovery. The JASPAR web interface allows users to browse TF binding profiles, search by identifiers or annotations, compare user-entered profiles with existing ones, and search nucleotide sequences against selected TF profiles. Selected profiles are displayed using sequence logos, with annotations linked to pop-up windows. JASPAR is tightly integrated with the TFBS Perl framework for TFBS analysis, providing a full-featured application programming interface. The TFBS framework has been enhanced to support word-based motifs and a leading word-based pattern discovery algorithm. The database is freely accessible and can be downloaded as a MySQL database dump or in flat-file format. JASPAR offers significant advantages over similar resources, including a non-redundant collection of reliable binding profiles, unrestricted data access, and strong integration with programming tools. It is distinct from TRANSFAC, which is a commercial resource with a broader mission and redundant profiles. JASPAR will continue to expand and improve existing profiles as new data emerge. Additional efforts are underway to incorporate yeast binding profiles and to integrate JASPAR with other collections for seamless programming-level integration.JASPAR is an open-access database containing high-quality, matrix-based transcription factor (TF) binding site profiles for multicellular eukaryotes. The profiles are derived from experimentally validated nucleotide sequences that bind TFs. The database includes a web interface for browsing, searching, and subset selection, an online sequence analysis tool, and programming tools for genome-wide and comparative genomic analysis. JASPAR is available at http://jaspar.cgb.ki.se. The database contains 111 profiles, derived from published collections of experimentally defined TF binding sites. Profiles were curated based on quality assessment criteria, including experimental validation, absence of arbitrary filtering, unambiguous sequence availability, and sufficient binding site sequences. Profiles were aligned using the ANNSpec program, which uses artificial neural networks and Gibbs sampling for pattern discovery. The JASPAR web interface allows users to browse TF binding profiles, search by identifiers or annotations, compare user-entered profiles with existing ones, and search nucleotide sequences against selected TF profiles. Selected profiles are displayed using sequence logos, with annotations linked to pop-up windows. JASPAR is tightly integrated with the TFBS Perl framework for TFBS analysis, providing a full-featured application programming interface. The TFBS framework has been enhanced to support word-based motifs and a leading word-based pattern discovery algorithm. The database is freely accessible and can be downloaded as a MySQL database dump or in flat-file format. JASPAR offers significant advantages over similar resources, including a non-redundant collection of reliable binding profiles, unrestricted data access, and strong integration with programming tools. It is distinct from TRANSFAC, which is a commercial resource with a broader mission and redundant profiles. JASPAR will continue to expand and improve existing profiles as new data emerge. Additional efforts are underway to incorporate yeast binding profiles and to integrate JASPAR with other collections for seamless programming-level integration.