Mobile crowd sensing (MCS) makes full use of the sensing and wireless communication capabilities of smart mobile devices to collect real-time information effectively. It makes it possible to monitor people’s health condition in real time. Our health information collected through MCS can be used to improve healthcare service. Hypertension is a widespread chronic disease, and preventing hypertension can effectively reduce the incidence of cardiovascular disease. In this paper, we propose a hypertension risk assessment approach based on mobile crowd sensing, which allows for real time health monitoring and warning. In order to stimulate the enthusiasm of MCS volunteers, optimized communication model is used to reduce the communication cost of non-data-users. Additionally, the current hypertension risk status of patients will be feed back to them in real time. In our approach, binary logistic regression is used to select risk factors of hypertension, and then the risk factors are used as the inputs of BP neural network to construct the risk prediction model. Furthermore, the hypertension risk is further divided into low risk, medium risk and high risk through cumulative distribution function. 4498 samples from a community health service center in Hefei area were used to evaluate the performance of the proposed approach. The experimental results show that the proposed approach can provide real-time, effective monitoring and dynamic feedback of the hypertension risk, offering a novel clinical tool for the early warning of hypertension. The proposed approach also provides a general framework for risk assessment of other chronic diseases.