Energetic use of plant oil becomes more attractive since sources of fossil fuels are expiring and therefore there prices rising. Especially the oil from the Jatropha curcas seeds becomes more attention since the plant comprise manifold applications. For the use of plant oil as household cooking fuel a plant oil pressure stove was developed. Nevertheless, the combustion of crude plant oil in pressure stoves leads to deposit formation inside the vaporizer. Therefore, the objective of this study was to establish a model for estimation of deposit formation from Jatropha oil quality parameters. Twenty different qualities of Jatropha oil were applied for combustion tests. Characteristic and variable properties were analyzed according standard methods. Model for prediction of deposits was developed by using a multiple linear regression analysis. Validation of the best fitting model was done by applying it for different types and qualities of plant oil. A highly significant adjusted R(adj.)(2) of 0.785 (p < 0.000) for the best fitting model indicated that deposits could be estimated by the parameters acid value, water content and ash content of J. curcas oil with satisfying accuracy. The coefficient of regression for validation of the model was R(2) = 0.756 and therefore acceptable. Following threshold values can be recommended for safe operation: acid value lower than 6.00 mg KOH/g, water content less than 0.15% and ash content below 0.10%. If those recommendations will be followed an undisturbed and reliable operation of plant oil pressure stoves with low maintenance requirements can be achieved. (C) 2010 Elsevier Ltd. All rights reserved.