miRecords is an integrated resource for microRNA (miRNA)-target interactions. It consists of two main components: Validated Targets and Predicted Targets. The Validated Targets component contains a large, high-quality database of experimentally validated miRNA-target interactions, with systematic documentation of experimental support for each interaction. It includes 1135 records of validated miRNA-target interactions between 301 miRNAs and 902 target genes in seven animal species. The Predicted Targets component stores predicted miRNA targets generated by 11 established miRNA target prediction programs. miRecords is expected to serve as a useful resource for both experimental miRNA researchers and informatics scientists developing next-generation miRNA target prediction programs. The database is available at http://miRecords.umn.edu/miRecords. miRNAs are small non-coding RNAs that regulate gene expression by base-pairing with target mRNAs, leading to downregulation or repression of the target genes. miRNAs are important in higher eukaryotes, with over 700 mature miRNAs identified in the human genome. The mechanisms by which miRNAs regulate their target genes are not fully understood, but several models suggest that miRNAs can induce translation repression or lead to rapid proteolysis of nascent polypeptides or accumulation of target mRNAs in P-bodies. Computational prediction of miRNA targets is more challenging in animals than in plants due to imperfect base-pairing between miRNAs and their target sites. Several miRNA target prediction programs have been developed, including TargetScan, PicTar, miRanda, DIANA-microT, RNAhybrid, MicroInspector, and Rna22. These programs use various methods, including hand-derived rules, sequence motifs, and machine learning techniques. However, no single program is consistently superior to all others, and experimental researchers often consider the intersections of predictions from multiple programs. miRecords addresses key issues in documenting validated miRNA-target interactions, such as distinguishing between endogenous and exogenous miRNA experiments, and classifying experimental evidence at the target gene, target region, and target site levels. It also includes detailed information on experimental methods, results, and target site mutations. The database provides a comprehensive collection of miRNA-target interactions, with a structured documentation system that differentiates between various types of experimental evidence. The miRecords web interface allows users to search for miRNA-target interactions, view predicted targets from multiple programs, and download validated targets. The database is publicly accessible and can be used for both research and development of miRNA target prediction programs. It is compared with other miRNA target resources, such as TarBase, and is noted for its comprehensive collection of miRNA-target interactions and structured documentation of experimental support.miRecords is an integrated resource for microRNA (miRNA)-target interactions. It consists of two main components: Validated Targets and Predicted Targets. The Validated Targets component contains a large, high-quality database of experimentally validated miRNA-target interactions, with systematic documentation of experimental support for each interaction. It includes 1135 records of validated miRNA-target interactions between 301 miRNAs and 902 target genes in seven animal species. The Predicted Targets component stores predicted miRNA targets generated by 11 established miRNA target prediction programs. miRecords is expected to serve as a useful resource for both experimental miRNA researchers and informatics scientists developing next-generation miRNA target prediction programs. The database is available at http://miRecords.umn.edu/miRecords. miRNAs are small non-coding RNAs that regulate gene expression by base-pairing with target mRNAs, leading to downregulation or repression of the target genes. miRNAs are important in higher eukaryotes, with over 700 mature miRNAs identified in the human genome. The mechanisms by which miRNAs regulate their target genes are not fully understood, but several models suggest that miRNAs can induce translation repression or lead to rapid proteolysis of nascent polypeptides or accumulation of target mRNAs in P-bodies. Computational prediction of miRNA targets is more challenging in animals than in plants due to imperfect base-pairing between miRNAs and their target sites. Several miRNA target prediction programs have been developed, including TargetScan, PicTar, miRanda, DIANA-microT, RNAhybrid, MicroInspector, and Rna22. These programs use various methods, including hand-derived rules, sequence motifs, and machine learning techniques. However, no single program is consistently superior to all others, and experimental researchers often consider the intersections of predictions from multiple programs. miRecords addresses key issues in documenting validated miRNA-target interactions, such as distinguishing between endogenous and exogenous miRNA experiments, and classifying experimental evidence at the target gene, target region, and target site levels. It also includes detailed information on experimental methods, results, and target site mutations. The database provides a comprehensive collection of miRNA-target interactions, with a structured documentation system that differentiates between various types of experimental evidence. The miRecords web interface allows users to search for miRNA-target interactions, view predicted targets from multiple programs, and download validated targets. The database is publicly accessible and can be used for both research and development of miRNA target prediction programs. It is compared with other miRNA target resources, such as TarBase, and is noted for its comprehensive collection of miRNA-target interactions and structured documentation of experimental support.