The identification of biomarkers for the early diagnosis of nonalcoholic fatty liver disease (NAFLD) is urgently needed. Here, we aimed to identify NAFLD biomarkers in the early stages of steatosis (SS) and nonalcoholic steatohepatitis (NASH) based on differential gene expression from bioinformatics data. A meta-analysis was performed from transcriptomic databases retrieved from public repositories containing data from biopsies of patients at various stages of NAFLD development. The status of the selected molecules was validated in the serum of patients with NAFLD by ELISA. : We identified 121 differentially expressed genes (DEGs) associated with SS and 402 associated with NASH. Gene Ontology (GO) enrichment revealed that the altered genes were primarily associated with dysfunction of primary cellular processes, and pathway analyses were mainly related to cholesterol metabolism. We identified ACSS2, PCSK9, and CYP7A1 as candidate biomarkers for SS and ANGPTL3, CD36, CYP51A1, FASN, FAS, FDFT1, and LSS as candidate biomarkers for NASH. By experimental validation of bioinformatics data from patients with NAFLD, we identified promising biomarkers for detecting SS and NASH that might be useful for screening and diagnosing early NAFLD stages in humans.