Health state assessment (HSA) which integrates structural and functional information of unit under test (UUT) can provide data support for maintenance decision of Conditon Based Maintenance (CBM). HSA technology based on health state-general test (HSGT) dependency is proposed in this paper. Firstly, according to different maintenance requirements, the degrees of fault severity are divided into several health states and the test ranges which are discretized by several general tests, thus a HSGT matrix can be built by integrating structural and functional information of a system. Secondly, the Bayesian theory is used to model HSA based on HSGT. Finally, the Monte Carlo simulation method is developed to build an eight-UUT system, and the effectiveness and feasibility of the proposed method is verified. The experimental results show that the proposed HSA technique can assess health states of UUTs in a timely and accurately way, and the assessment results can trigger the maintenance decision prior to functional failures due to fault growth.