Objective To analyze and screen the hub genes and signal pathways related to the recurrence of liver cancer based on bioinformatics, so as to provide a theoretical basis for exploring the molecular mechanism of liver cancer recurrence. Methods Datasets related to liver cancer were downloaded from The Cancer Genome Atlas (TCGA) database and Gene Expression Omnibus (GEO) database, and differential expression genes were screened by edgeR algorithm, and then they were evaluated by Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). Protein-protein interaction (PPI) network was constructed to determine the hub genes related to liver cancer recurrence. After the hub genes were selected, least absolute shrinkage and selection operator (LASSO)-logistic regression analysis was used to construct the recurrence prediction model. Single sample gene set enrichment analysis (ssGSEA), estimation of stromal and immune cells in malignant tumour tissues using expression data (ESTIMATE) and microenvironment cell populations-counter (MCP) algorithms were used to evaluate immune infiltrating cells. Results A total of 343 recurrence related differentially expressed genes were screened. Go and KEGG enrichment analyses indicated that most of the genes were enriched in biological processes such as forebrain development, regulation of chemical synaptic transmission and signaling pathways such as neuroactive ligand-receptor interaction. PPI network analysis screened 12 hub genes. An independent risk-score model was determined by using LASSO-logistic regression analysis based on 12 hub genes, with a calculation formula:risk score=22.1-somatostatin (SST) expression value×0.21-dopamine receptor D2 (DRD2) expression value×1.94-calbindin 2 (CALB2) expression value×1.153. The risk model showed a good predictive ability with an area under curve of 0.683. Moreover, it was proved that CALB2 was lowly expressed in tumor tissue and positively correlated with immune infiltrating cells in tumor microenvironment of recurrent liver cancer. Conclusion The recurrence related gene model can provide better efficacy to predict the recurrence of patients, and the hub gene CALB2 plays a crucial role in the recurrence of liver cancer.