Analysing economic efficiency of farm production always faces a problem of insufficient information. This is particularly true when the analysis is performed on the reference farm where estimates are based on the average aggregated data. The paper illustrates how the combination of different mathematical programming methods could be efficiently used to analyse the farm-production plan with the lack of the on-farm accounting data. The utilised approach shows how the holistic analysis of production planning as a multi-criteria problem could be conducted. The estimation of the missing information and the disaggregation of the endogenous farm data is enabled through different models that are based on the constrained optimisation. The developed models are linked into the spreadsheet modular tool enabling systematic analyses of the farm decision making under risky conditions. Illustration of the modular tool application is given via the analyses of three hypothetic dairy farms. The obtained results indicate that the developed approach enables holistic analyses of the production planning. The methodology applied provides also important information for the measures aimed to increase efficiency as well as to benchmarking the performance of different farm types. The results point to a discrepancy between the solutions obtained through different objective functions and shows the advantage of the multi-criteria approach.