In order to solve the problem of high power consumption and limited node energy in the existing aquaculture water quality monitoring system, a low-power aquaculture water quality monitoring system is designed. For wireless testing nodes, cycle adaptive adjustment algorithm is used to reduce the number of data transmission and power adaptive adjustment algorithm is used to reduce power consumption. For the testing gateway, NB-IoT technology is mainly used to reduce wireless transmission power consumption and FreeRTOS real-time operating system is carried to increase the CPU utilization to reduce power consumption. Experiments show that under the condition that significant change data is not lost, the number of data transmission is reduced by 140% and 20% respectively for using cycle adaptive adjustment algorithm and using fixed cycle 15 min and 30 min. Under the condition that packet loss required by the system is satisfied, power consumption for using the node of power adaptive adjustment algorithm is 9.5%-42.5% less than that for using the node of fixed transmit power 4 dBm. Experimental results show that the system has the characteristics of low power consumption, high stability, sensitive remote control, and has a high application value.