Hepatocellular carcinoma (HCC) is one of the most common malignant tumors worldwide, and current treatments exhibit limited efficacy against advanced HCC. The majority of cancer-related deaths are caused by metastasis from the primary tumor, which indicates the importance of identifying clinical biomarkers for predicting metastasis and indicating prognosis. Patient-derived cells (PDCs) may be effective models for biomarker identification. In the present study, a wound healing assay was used to obtain 10 fast-migrated and 10 slow-migrated PDC cultures from 36 HCC samples. MicroRNA (miRNA) signatures in PDCs and PDC-derived exosomes were profiled by microRNA-sequencing. Differentially expressed miRNAs between the low- and fast-migrated groups were identified and further validated in 372 HCC profiles from The Cancer Genome Atlas (TCGA). Six exosomal miRNAs were identified to be differentially expressed between the two groups. In the fast-migrated group, five miRNAs (miR-140-3p, miR-30d-5p, miR-29b-3p, miR-130b-3p and miR-330-5p) were downregulated, and one miRNA (miR-296-3p) was upregulated compared with the slow-migrated group. Pathway analysis demonstrated that the target genes of the differentially expressed miRNAs were significantly enriched in the 'focal adhesion' pathway, which is consistent with the roles of these miRNAs in tumor metastasis. Three miRNAs, miR-30d, miR-140 and miR-29b, were significantly associated with patient survival. These findings indicated that these exosomal miRNAs may be candidate biomarkers for predicting HCC cell migration and prognosis and may guide the treatment of advanced HCC.