Monitoring of human motor and muscle activity is used in many areas, from prosthetics during rehabilitation to training monitoring of athletes. Sensors for these tasks are usually made of flexible polymers and require recycling after the expiration date. Nanocellulose (NC) can be used as a biodegradable base for this type of sensor. The development of low-cost disposable sensors that do not require disinfection and cleaning is relevant. NC is a composite nanoscale structure of cellulose fibers (fibrils) with a high aspect ratio. The paper aim is to develop disposable wearable biodegradable bend sensors based on nanocellulose using vacuum synthesis methods and the study of their characteristics. Nanocellulose was synthesized by the TEMPO method. The sensors were created by means of magnetron sputtering of Ti/Ni or Cr/Ni thin films at the surface of nanocellulose. Measuring stand was developed to determine the change in resistance due to the bending of the sensor. It’s mechanical part consists of an elastic deformation plate made of high-alloy steel, which can be bent using a micrometric screw. The change in resistance is linearly related to the elongation of the measured sample. A Wheatstone bridge and a 24-Bit ADC HX711 were used to measure the change in resistance. During testing of the sensor for analysis of muscle activity, the sensor element was attached to the human skin with the help of medical glue BF-6. The obtained sensors were tested for biodegradability. The samples were placed in the ground at a depth of 20-30 mm. The mass of nanocellulose samples was measured using a high-precision digital balance EDIS 50 (50/0.001 g) with a built-in level. The optimal ratio of the value of sensitivity and reversibility is observed in the range of the nominal resistance of the nickel film from 10 to 100 Ohms. This is due to an increase in the surface area of the Ni film, which leads to an increase in sensitivity, but at the same time there is a decrease in the repeatability of the characteristics due to a greater influence of the heterogeneous structure of nanofibrillated cellulose. In addition, sensors with different buffer layer materials - Ti and Cr - were selected for testing. For titanium-based sensors, the maximum sensitivity coefficient is 0.312%, while the deviation of the sensor signal after one bending-unfolding cycle (reversibility) is less than 0.001%. Chromium-based sensors have significantly higher sensitivity (0.9753%), but worse reversibility (7.14%). Sensors based on the Cr buffer layer showed poorly reproducible results in the cyclic mode of operation, namely: there are significant fluctuations in the signal amplitude (up to 50-60%) already after the second bending-unfolding cycle. Therefore, despite the high sensitivity of such sensors, they are unsuitable for analyzing human motor and muscle activity/ The sensors based on the Ti buffer layer showed good response (2.5-3%) and good repeatability and resistance to cyclic bending (30 times). It can be seen that the obtained dependencies are approximated by a linear law. Some deviation from linearity is obviously related to the inhomogeneity of the Ni thin film. Also, the sensors showed a good loss of mass (40% in 9 weeks) during the biodegradability test, which confirms their ability to decompose under the influence of atmospheric phenomena. So, in this work, disposable wearable sensors on a nanocellulose substrate were synthesized for the evaluation of motor and muscle activity of a person. It was found that such sensors can be used to test of finger and biceps movement during at least 10-30 full flexion-extension cycles. For test of elbow movement, it is planned to synthesize a high-elastic composite material based on nanocellulose and bioelastic material (for example, polyvinyl alcohol). Thus, the proposed sensor manufacturing technology makes it possible to obtain cheap, light, flexible disposable wearable sensors that do not require further disposal after the end of operation.