The role of cancer-associated fibroblasts (CAFs) has been thoroughly investigated in tumour microenvironments but not in bladder urothelial carcinoma (BLCA). The cell fraction of CAFs gradually increased with BLCA progression. Weighted gene coexpression network analysis (WGCNA) revealed a specific gene expression module of CAFs that are relevant to cancer progression and survival status. Fifteen key genes of the module were consistent with a fibroblast signature in single-cell RNA sequencing, functionally related to the extracellular matrix, and significant in survival analysis and tumour staging. A comparison of the luminal-infiltrated versus luminal-papillary subtypes and fibroblast versus urothelial carcinoma cell lines and immunohistochemical data analysis demonstrated that the key genes were specifically expressed in CAFs. Moreover, these genes are highly correlated with previously reported CAF markers. In summary, CAFs play a major role in the progression of BLCA, and the 15 key genes act as BLCA-specific CAF markers and can predict CAF changes. WGCNA can, therefore, be used to sort CAF-specific gene set in cancer tissues. K E Y W O R D S bladder cancer, cancer-associated fibroblast, marker, tumour microenvironment, weighted gene co-expression network analysis 1 | INTRODUCTION Bladder urothelial carcinoma (BLCA) is one of the leading causes of cancer-related death worldwide, with a 5-year survival rate of only 5% in patients with metastases (Antoni et al., 2017). BLCA with local and distant metastases remains an urgent challenge for researchers. Most cancers exhibit epithelial abnormalities and mutations in transformed cells. Over the past 20 years, this scenario has evolved, and the matrix has been shown to act as a driver of the tumourigenic process and promote cancer progression (Gascard & Tlsty, 2016). Cancer-associated fibroblasts (CAFs) are the main components of the tumour microenvironment (TME), and they appear in the matrix surrounding cancer. Biochemical crosstalk between cancer cells and CAFs and the mechanical remodelling of the stromal extracellular matrix (ECM) by CAFs are important factors in tumour cell migration and invasion (Erdogan & Webb, 2017; Valkenburg, de Groot, & Pienta, 2018). However, the current understanding of the interplay between CAFs and TME in BLCA is limited. High-throughput data from a large number of patient samples reveal the pathogenesis of and mechanisms underlying various cancers. In combination with weighted gene co-expression network analysis (WGCNA), the search for highly co-expressed key genes or hub genes is also common in tumour progression (Giulietti et al.,