A large number of cases showed that fentanyl (FEN) has become the main cause of death from illegal drug overdose owing to its potent effect on respiratory depression, which has emerged as a grave threat to public health and safety. However, traditional analytical methods require cost-prohibitive equipment, complex pretreatment procedures, and technically trained experts, thus highlighting the urgent need to develop a cost-effective, straightforward, and highly sensitive method to detect FEN. This work demonstrated a dual-readout sensor FGGC-AuNCs@Q7 for FEN detection, which is based on the molecular recognition and self-assembly between the macrocycle cucurbit[7]uril (Q7) and FEN, accompanying spontaneous visual Tyndall effect and fluorescence optical responses of the gold nanoclusters within seconds. A detection limit of 1 ng mL -1 and a linear range of 9 to 148 000 ng mL -1 were achieved for fluorescence detection on FEN, with favorable selectivity in the presence of other illicit drugs or common interferents. The proposed method has been proved by its satisfactory application for the analysis of human urine.