A method for the analysis of the effects of subjective and objective tolerances in networks and systems is presented in this paper. The imprecision of the components is represented using fuzzy sets and then the value of the desired attribute is computed. The resulting attribute is also fuzzy and is obtained in the form of a fuzzy set. This fuzzy set contains the extremal and other values of the desired attribute along with their grades of membership. Thus, apart from getting the extremel values, we get the overall picture of the attribute. Since this method is directly applicable for the analysis of subjective tolerances, networks representing humanistic systems may be analysed using this method.