Wireless sensor networks (WSNs) implemented in the paradigm of the Internet of Things (IoT) are characterized by a large number of distributed sensor nodes that make measurements in-the-field and communicate with other sensor nodes and servers in the cloud by means of wireless technology. Sensor-to-microcontroller direct interface (SMDI) is a technique used for the measurement of resistive sensors without the use of an ADC. In SMDI based measurements, the sensor is directly interfaced with the digital input-output pins of the general purpose input output (GPIO) interface of microcontrollers and FPGAs. Compared with the measurements per-formed with an ADC, SMDI is characterized by lower cost and lower power consumption. In this paper, the impact of noise on the accuracy of resistive sensor measurements using SMDI is investigated. The study was carried out by LTSpice electrical level simulations and validated by preliminary experimental measurements, where a set of resistances in the range from 100 Ω to 10 kΩ were measured by SMDI under different levels of noise. For each operative condition, the simulations were also carried out in the case of measurements performed with a 12-bit ADC and the achieved accuracy for the measured resistances was compared with the results achieved by SMDI. The results have shown that noise can seriously impact the measured accuracy of resistive sensors by SMDI and, differently from the ADC measurements, the accuracy cannot be improved by averaging on multiple measurements. A mitigation strategy to estimate the noise level and to improve the measurement accuracy of resistive sensors by SMDI was also proposed.