It has recently been suggested that attribution theory expand its locus of causality dimension beyond internal and external attributions to include relational (i.e., interpersonal) attributions (Eberly, Holley, Johnson, & Mitchell, 2011). The current investigation was designed to empirically focus on relationship dynamics, specifically where 1 member of the relationship receives negative performance-related feedback. We use quantitative and qualitative data from 7 samples (5 samples for scale validation in Study 1 and 2 for hypothesis testing in Studies 2 and 3) that provide empirical support for the existence and impact of relational attributions. Our findings identify the circumstances under which relational attributions are likely to be formed and indicate that relational attributions are related to relational improvement behaviors, particularly when employees are of the same sex as their relationship partner and perceive sufficient time and energy to engage in relational improvement efforts. A personal attribute, relational self, contrary to expectations, did not moderate the link between relational attributions and improvement behaviors. These findings make an important contribution to attribution theory by providing evidence for the criticality of relational attributions within interdependent work relationships.