The purpose of this research is to investigate the use of dyads instead of whole rating groups for analyzing 360° feedback. A Center for Creative Leadership sample of direct subordinates was matched with their focal leaders (J = 4810 dyads). Two CCL Benchmark scales (Building & Mending Relationships and Problems with Interpersonal Relationships) were used to determine if this dyadic pairing produced high self–other agreement (SOA). Using Within and Between Analysis (WABA), results suggest the viability of detecting unitary, whole dyadic effects only under specific boundary conditions. Other conditions produced highly dispersed dyadic results. Thus, dyadic convergence is not a universal phenomenon in these data. Tying the dyadic results back to the original rater group's consistency as reported in Markham, Smith et al. (2014), those rater groups with high internal agreement form clearer dyadic pairings when compared to rater groups that were moderate or low in agreement. Implications and future directions are discussed.