This paper reports results of two studies which investigated elements of the current (House & Dessler, 1973; House & Mitchell, 1974) and initial (House, 1971) versions of the path-goal theory of leadership. The first constituted a replication and the second a longitudinal study which tested several assumptions about causation which underlie both versions of the theory. The findings largely supported the theory with the notable exception of the hypotheses concerning subordinate performance.