The methodologies presented in Parts I and II (refer Modak and Lohani, 1984a and b) are essentially for deciding the best number and configuration for a single pollutant monitor. In practical situations however, Air Quality Monitoring Networks (AQMN) are expected to measure more than one pollutant and therefore simultaneous consideration of different types of pollutants must be made.In this paper, two new approaches have been developed for the multipollutant AQMN design. The first method makes use of the index theory and the other makes use of the principles of Pareto optimality. As an illustration of these methodologies, an example from Taipei City, Taiwan is considered.