Bladder cancer is the 9th leading cause of cancer death worldwide. The major problem in bladder cancer is primarily the high recurrence rate after drug treatment and resection. Although conventional screening methods, such as cystoscopy, urinary cytology and ultrasound sonography, have become widely used in clinical settings, the diagnostic performance of these modalities is unsatisfactory due to low accuracy or high invasiveness. Because circulating micro RNA (miRNA) profiles have recently been reported as an attractive tool for liquid biopsy in cancer screening, here, we performed global miRNA profiling of 392 serum samples of bladder cancer patients with 100 non‐cancer samples and 480 samples of other types of cancer as controls. We randomly classified the bladder cancer and control samples into 2 cohorts, a training set (N = 486) and a validation set (N = 486). By comparing both controls, we identified specific miRNA, such as miR‐6087, for diagnosing bladder cancer in the training and validation sets. Furthermore, we found that a combination of 7 miRNA (7‐miRNA panel: miR‐6087, miR‐6724‐5p, miR‐3960, miR‐1343‐5p, miR‐1185‐1‐3p, miR‐6831‐5p and miR‐4695‐5p) could discriminate bladder cancer from non‐cancer and other types of tumors with the highest accuracy (AUC: .97; sensitivity: 95%; specificity: 87%). The diagnostic accuracy was high, regardless of the stage and grade of bladder cancer. Our data demonstrated that the 7‐miRNA panel could be a biomarker for the specific and early detection of bladder cancer. The present study selected 7 miRNA expression levels among bladder cancer, non‐cancer and other cancer samples. A diagnostic index was calculated and plotted in the dot plot among 12 different cancers. An index score ≥0 indicated the presence of bladder cancer and an index score <0 indicated the absence of bladder cancer.