In this paper, both coking and non-coking coals of India have been ranked for industrial use, using a powerful and sophisticated technique called fuzzy multi attribute decision making. Proximate analysis has been taken as the basis and the four important parameters of coal, viz. fixed carbon, volatile matter, moisture content and ash content have been considered for ranking purposes. These parameters have been treated as fuzzy sets over the ranges prescribed in literature. In order to give due weightages to the coal quality parameters depending on their importance, Saaty's analytic hierarchy process (AHP) (Saaty, 1983, 1992; Hanratty and Joseph, 1992) has been adopted. Later, Yager's (1978) fuzzy multi attribute decision making approach has been employed to arrive at the coal field having the best quality coal for industrial use. Also, Yager's model has been extended in that the three different aggregators have been used instead of the min operator. Three different kinds of membership functions, viz. linear, exponential and S-type for the parameters coupled with four kinds of aggregators: (i) min operator, (ii) fuzzy and, (iii) compensatory and (iv) a linear combination of min and max operators have been studied here. The S-type membership function, in combination with all the types of aggregators, turned out to be quite robust for the coking coals whereas for non-coking coals the S-type membership function with a linear combination of min and max operators proved to be superior to other combinations. The results, though, matched with an earlier study based on factor analysis in ranking the same field as the best quality field in both coking and non-coking coals, they differed significantly in the case of other coal fields. The advantage of the present approach is that the inherent fuzziness in the traditional treatment of the coal parameters has been modelled for the first time in a sophisticated and systematic mathematical framework. Also, the all important parameter, viz. fixed carbon has been given due weightage in this paper, which has not been done in earlier studies. These are the reasons behind the mismatch found between the present study and the one based on factor analysis.