In Fiedler's contingency theory, situation favourableness for leader depends upon three situational variables — Group Atmosphere, Task Structure and Position Power. Each variable is dichotomized as high or low to define a 2 × 2 × 2 factorial cube with eight octants, and correlations between leadership style and effectiveness are plotted against the octants. To place these octants along a one-dimensional scale of favourableness, Fiedler assumes that Group Atmosphere is most important and Position Power is least important, and that the situational variables combine according to an adding rule. In four judgmental experiments conducted during different periods of national emergency in India, favourableness of leadership situations was studied with the methods of Anderson's information integration theory. Results indicated (a) that relative importance of situational variables changed across situations, (b) that situational variables were averaged in judgment of situation favourableness, and (c) that spacing of octants on the horizontal axis according to their functional measurement values generated a considerably better bow-shaped curve for correlation between leadership style and effectiveness than was obtained with Fiedler's octant scale. Judgmental experiments seemed to have great potential for providing a more analytic approach to further work on contingency theory and leadership behaviour.