The dataset represents transcriptomes of liver (LI), myocardium (HM), and skeletal muscle (SM) from 41 mostly preterm human newborns who mainly died within several hours or days after birth. The Excel files contain raw read counts of all detected genes before filtering and filtered protein-coding genes data [raw, normalized, and regularized logarithm (log) transformed read counts excluding low-expressed genes]. Supported by the Ministry of Health of the Czech Republic, grant no. NU20-07-00026 and by the project National Institute for Research of Metabolic and Cardiovascular Diseases (Programme EXCELES, ID Project No. LX22NPO5104) funded by the European Union - Next Generation EU. The data provide a unique resource for the research of human perinatal development. The general description and analyses of all the data in all three tissues can be found in Janovska et al, 2025, Frontiers in Cell and Developmental Biology (doi: 10.3389/fcell.2025.1645959). A subset of the data has been used to characterize hepatic haematopoiesis in human newborns, as described in Janovska et al, 2025, Frontiers in Pediatrics (doi: 10.3389/fped.2025.1572836).