A multitude of eco-design methodologies applicable to the scale of the product exists currently. The majority of these approaches use life cycle assessment (LCA). These approaches are generally based on comparisons of preset scenarios in order to choose the most favorable scenario from the environmental point of view. Various variability sources can be present during the environmental evaluation and will not be taken into account using this approach. We propose in this article to characterize and quantify the effects of these variability sources on the LCA results. We thus study the share of influence of three types of parameters: (1) technological, (2) environmental and (3) methodological on selected indicators of the CML2001 characterization method. The following indicators were used: GWP100, AP, EP, ETP100 representing all indicators of ecotoxicity, HTTP100 and energy demand ED. During our application on the hemp crop we found a potential reduction of environmental impact from 30 to 88% using technological option, and we identify also those models which need to be improved.