During colorectal carcinogenesis, a spectrum of lesions is formed with different malignant potentials and occasionally with ambiguous morphologic features. To identify novel biomarkers, we previously performed a bioinformatics study to detect differentially expressed genes between colorectal adenoma (CRA) and colorectal carcinoma (CRC). The present study validated the diagnostic and prognostic significance of 6 components of the extracellular matrix (ECM) that were identified in the previous bioinformatics analysis [decorin (DCN), erythropoietin‑producing hepatoma receptor A4 (EPHA4), fibronectin 1 (FN1), secreted protein acidic and cysteine rich (SPARC), spondin 2 (SPON2) and secreted phosphoprotein 1 (SPP1)], by analyzing their gene and protein expression levels in all samples. A total of 60 formalin‑fixed paraffin‑embedded biopsy samples were included in the present study, from 40 patients with CRA, malignant polyps and CRC with and without lymph node metastases. The expression of the ECM‑related genes was evaluated on the mRNA level using reverse transcription‑quantitative PCR (RT‑qPCR) and on the protein level using immunohistochemistry. RT‑qPCR results revealed differential expression among the groups for all genes, except for EPHA4. Their expression proportionally increased with the level of malignancy. Among them, SPARC, SPON2 and SPP1 were differentially expressed between CRA and malignant polyps. Immunohistochemistry analysis revealed two patterns: Positive staining of SPON2 and SPP1 in epithelial cells of healthy mucosa and dysplastic glands of CRA and CRC, and positive staining of DCN and SPARC in the lamina propria in healthy mucosa and stroma of CRA and CRC. The intensity of staining increased with the severity of lesions. The present results suggested that ECM‑related proteins may have an important role in the development of CRC, with a possible diagnostic use for differentiation among various lesions, such as between CRA and malignant polyps. However, no significant potential was detected for these genes to predict lymph node metastases in CRC.