Researchers have employed two alternative approaches in measuring the sources of marketing channel power: (1) an indirect assessment through assistances and punishments and (2) a direct measurement of each specific power source. This study compares empirically the construct validity of both approaches. The results indicate that the reliability, the content validity, the within-method and across-method convergent validity, and the discriminant validity of both approaches are acceptable. While the a priori dimensionality of the direct approach was recovered, a dimension not hypothesized was found for the indirect approach to power source measurement. Because the direct power source measures were significantly related to the attributions of power while the indirect ones were not, the direct approach demonstrated better nomological validity. We were not, however, able to assess the impact of shared methods variance upon the validity of these measurement approaches. Overall, the direct approach to measuring the sources of marketing channel power appears to have greater construct validity.