The paper has solved the problem of optimizing the existing inefficient and irrational system of the quantitative monitoring of the situation in peacetime fuel supplies at the pumping stations in the Army of Serbia. A study of existing organizational forms, military pumping stations as well as civilian ones, was carried out. Based on the completion of the survey by competent persons in the military, the methods of expert evaluation and the obtained quantitative indicator of the tested models, a multicriteria optimization was performed in order to select the optimal model. The optimization of the existing models, in terms of efficiency and economy, would be the rationalization and modernization - automation of military capacity and greater reliance on automated civilian pumping stations. Introduction Within the framework of the undergoing reform of the Serbian Army and in order to reduce the total costs, it is necessary to optimize the existing supply system that is technologically outdated, inefficient and uneconomic. The problem of research in this paper is reduced to the selection of an optimal model of the quantitative monitoring of the state of peacetime stocks of fuel at the pumping stations in the Serbian Army, in order to ensure economical operation and efficient monitoring of available and issued quantities, aiming at better decision making and management in the supply system as well as at achieving faster system response, with greater reliance on government logistics. Organization of work and monitoring the fuel quantitative status at pumping stations The existing system of monitoring the quantitative state of fuel pumping stations in the Army of Serbia has the following disadvantages: getting unreliable data, due to outdated equipment for fuel handling and measuring equipment, and manual collection of data; creation of unauthorized shortages (due to subjective human error or deception); inadequate engagement of respective material and human resources. Optimization of quantitative monitoring of peacetime supplies of fuel at gas stations should aim at reducing the impact of the human factor, introducing automated quantitative monitoring of fuel condition with modern equipment for handling as well as applying technology for fast reading and dissemination of information and reports. Civilian pumping stations have been modernized gradually with new digital pump machines, systems for automated production and automated systems for measuring the fuel level in buried tanks. The objectives and criteria of the optimization of model monitoring In order to solve the problem of multi-criteria nature, the methods of operational research have been applied and the formalization of problem solving has been carried out. Models have been identified, criteria and subcriteria have been defined as well as respective criteria values, sub-criteria and weight coefficients for chosen variants in order to rank the alternatives - models. On the basis of the defined objectives and optimization approaches, the task of optimization to be solved is to choose one optimal model of monitoring the quantitative condition of peacetime stocks of fuels at gas stations, out of three variations or alternative models. Application of expert assessment and methods of analytical hierarchy process The problem was solved first 'manually', by using MS Excell, and after that by using the Expert Choice software package. The Expert Choice software package is based on the application of the method of analytical hierarchy process and combines the benefits that this method offers with the speed and visibility of computerized calculations and their result display. The purpose of the AHP method is to rank alternative decisions by their importance and to select the most acceptable alternative on the basis of a defined set of criteria and alternatives. The problem of determining the weight of criteria has been determined by applying the method of expert evaluation, along with interviewing and taking into account the competence of the interviewees. Conclusion The existing solutions in the system of logistics support, focused on the monitoring of the quantitative status of peacetime fuel supplies at the pumping stations in the Serbian Army, are not the product of system analysis and optimization. Therefore, it is not rational for them to get an IT support. The continuous task should thus be to monitor the phenomena and problems and their formalized description in order to create a foundation for their automation and successful problem solving.