Nowadays radars of air surveillance system are the main air traffic control system for the airspace around airports. They are applied for detection object’s parameters, such as location, position, coordinates, angle, size, speed, and others. But a lot of airport surface applies multiple radar coverage, that’s why fusion of information from different radars is one of the necessary procedures to provide a reliable object coordinate determination. The surveillance system employs data fusion algorithms to combine measurement data from radars for object coordinate determination and minimization of measurement error. Therefore, this paper focuses on researching the use of Dempster-Shafer evidence theory to combine surveillance data in order to process imprecise, inconsistent and partial data, received from various surveillance radars for object coordinate determination. The developed approach applies new algorithm of calculation of basic probabilities and mathematical expectation of coordinates of the object, integrating initial data of several surveillance radars and characteristics of their zones of ambiguity. The scientific novelty of this paper is the proposition of new function – basic probability density of basic probability (basic mass) for Dempster-Shafer evidence theory. Applying this new approach, we can determine most likely coordinates of the object. It also was considered three numerical examples of object coordinate determination applying evidence theory. This approach can be applied for solving various civil and military problems.