Cryogenic waterless live transportation (CWLT) is gaining attention as an efficient and green transportation strategy. However, how to non-destructively assess the stress level of live fish and thus improve the transportation management is a challenging goal. In this study, a highly integrated and low-cost live fish wearable electronic system (LFWES) was developed and applied. The system consists of a PDMS-based pneumatic direct written stretchable circuit (SC) and an fPCB. The good structural design ensures the advantage of small size and flexibility, while multimodal wireless sensing of temperature, humidity, and Euler angles can still be achieved after cyclic stretching. We proposed for the first time to wear SC on the gill cover for non-destructive monitoring of respiratory state as well as microenvironmental parameters, and thereby characterizing physiological stress. During the 10 h simulated CWLT of sturgeon, the patterns of change in stress state were obtained for acute stress, stress adaptation, and stress accumulation, which were in good agreement with the blood glucose concentration. Meanwhile, the fuzzy neural network (FNN) was employed to fuse modeling for environmental and physiological factors, and stress level assessment accuracy reached 88.1%. The results indicate a step toward the development of intelligent sensing capability in fisheries.