Traditional Chinese Medicine treats pulse diagnosis as a very important diagnostic tool which can give very detailed information on the state of the state of internal organs. However, it is seldom used by people in monitoring their own daily health because it requires training and experience. We've designed a personal 'Pulse Reading' sensor gadget that can take the pulse information from three different wrist positions at the same time and detect the frequency, stress, rhythm and the change under three different pressures. It can also communicate with a personal computer and show readings that can be used for stress recognition based on Traditional Chinese Medicine. A wide market survey of this kind of product was done to set the value of the product before starting to make the prototype. The MEMS sensor and sensor package survey were studied in order to build an accurate and stable sensor of right size and with appropriate output and a reasonable price. Pulse signals were denoised and screened through matlab and stress pattern recognition was built using a BP neural network. Finally, we built the prototype which consists of a pulse reading gadget, an analog/digital conversion part which lets the sensor gadget communicate with a computer and the matlab programming, which process the pulse signals and identifies the stress levels. The accuracy of the testing data for classifying stress is about 75%. The product cost is 100 times lower than similar existing commodity. Overall, we believe that this product will let people monitor their health and benefit from this information more conveniently and effectively than the can at present.