Understanding the impact of clinically-relevant variants is dependent on accurate and comprehensive annotation of genes and transcripts. However, the lack of standardised “default” transcripts for clinical variant reporting and for display across resources can present challenges. To promote harmonisation, EMBL-EBI and NCBI, the two providers of comprehensive human transcript annotation (Ensembl/GENCODE and RefSeq), are working together on the Matched Annotation from the NCBI and EMBL-EBI (MANE) project. We aim to generate a genome-wide transcript set that: 1) matches GRCh38, 2) includes pairs of 100% identical RefSeq (NM) and Ensembl/GENCODE (ENST) transcripts, thus allowing synonymous use of identifiers and 3) is high-confidence based on conservation, expression and overall support. The MANE Select set includes one transcript at each protein-coding locus across the genome that is representative of biology at that locus. To select transcripts, each group followed a computational approach, taking into account evidence of functional potential such as transcript expression, evolutionary conservation, and clinical significance. Curators from both groups provide a layer of manual review to ensure that automated selections are ideal or to make transcript decisions for complex loci. The MANE dataset is versioned and largely stable, only allowing updates if absolutely required. Currently, we have defined a MANE Select for 87% of coding loci, including all ACMG59 genes and over 90% of known clinical genes. We have also introduced the MANE Clinical set to support clinical reporting in complex genes that require more than one transcript to represent all clinical variants. Released transcripts are available on all genome browsers, and in commonly accessed resources such as gnomAD, dbSNP, and ClinVar. We encourage feedback on the MANE Select and Clinical sets to ensure that clinical and research needs are met (mane-help@ebi.ac.uk).