Radio frequency identification (RFID) tags are widely used in various electronic devices due to their low cost, simple structure, and convenient data reading. This topic aims to study the key technologies of ultra-high frequency (UHF) RFID tags and high-precision temperature sensors, and how to reduce the power consumption of the temperature sensor and the overall circuits while maintaining minimal loss of performance. Combined with the biomedicine, an innovative high-precision human UHF RFID chip for body temperature monitoring is designed. In this study, a ring oscillator whose output frequency is linearly related to temperature is designed and proposed as a temperature-sensing circuit by innovatively combining auxiliary calibration technology. Then, a binary counter is used to count the pulses, and the temperature is ultimately calculated. We designed a relaxation oscillator independent of voltage and current. The various types of resistors were used to offset the temperature deviation. A current mirror array calibration circuit is used to calibrate the process angle deviation of the clock circuit with a self-calibration algorithm. This study mainly contributes to reducing power consumption and improving accuracy. The total power consumption of the RF/analog front-end and temperature sensor is 7.65 μW. The measurement error of the temperature sensor in the range of 0-60 ℃ is less than ±0.1%, and the accuracy of the output frequency of the clock circuit is ±2.5%.