Lameness is, after mastitis and reduced fertility, the third most important health problem in dairy cows. Lameness causes a significant decrease in milk production, a reduced fertility, a higher culling rate and severely deteriorates animal welfare. To help farmers in detecting lame cows, ILVO has developed the GAITWISE system which consists of a six-meter long pressure sensitive mat to detect the positions of the legs with respect to time as well as the relative pressure exerted by the legs when cows walk over it. In order to determine whether a cow is lame, measured variables (e.g. stride length, abduction, asymmetry in relative force) are compared with the previous values of the same cow by a cow specific model. To validate the measurements from the GAITWISE, an open-source, low-cost and adaptable prototype 3D accelerometer is developed. The goal is to detect the motion of the cow’s legs using accelerometers and gyroscopes in x, y and z-direction in a way that data can be obtained online in order to reconstruct the motion of a cow’s legs. For the first prototype the single-board microcontroller Arduino Nano with an Atmega 328, an open source platform, serves as the processing unit, while MPU-6050 (InvenSense, CA, USA) is used to get the raw tri-axis acceleration and angular velocity data. The MPU-6050 is designed for low power, low cost and high performance requirements of wearable sensors. The unit has a digital output and communicates with the Arduino platform through I2C. An ultra-low power module (NRF24L01+, Nordic Semiconductor, Trondheim, Norway) is used to create a network that allows to receive the data online using the 2.4 GHz ISM band. The newly developed sensor is user programmable, allowing one to choose between different ranges and resolutions for the measurement of the acceleration (±2g, ±4g, ±8g or ±16g) and the angular velocity (±250°/s, ±500°/s, ±1000°/s or ± 2000°/s). An Arduino Nano with a high power NRF24L01+ module and a rechargeable battery is placed on the neckband of the animal as an extra node in the network to amplify the signal. The receiver module contains also an Arduino Nano and a high power NRF24L01+ module and is connected to a computer to collect and process all the received data. The price for the sensor on the legs (one per leg) will be approximately € 20.5, the neckband transmitter and the receiver around € 26 and € 17 respectively, which adds up to a total of € 125. First tests showed that the prototype is able to measure accelerations with an error below 2.5 % and gyroscope values with an error below 0.3 %. A preliminary test with a walking cow showed that the sensor was able to record the acceleration and angular velocity of the cow’s legs. Moreover, the designed sensor is open source and low-cost. The main drawback of our first prototype is that after 24hrs the battery needs recharging. To be used in future tests with walking cows a smaller version of the sensor that is attached to the legs, will be designed. Therefore the Pro Mini version of the Arduino microcontroller will be used with a NRF24L01+ low power transceiver and a rechargeable battery.