Most work-family conflict (WFC) research does not theorize, hypothesize, or empirically test phenomena at the dimension level. Instead, researchers have predominantly used composite-level approaches based on the directions of WFC (work-to-family and family-to-work conflict). However, conceptualizing and operationalizing WFC at the composite level instead of at the dimension level has not been confirmed as a well-founded strategy. The goal of the current research is to explore whether there is theoretical and empirical evidence in the WFC literature to support the importance of dimension-level theorizing and operationalization when compared to composite-level approaches. To advance theory related to the dimensions of WFC, we begin by reviewing WFC theories and then demonstrate the relevance of resource allocation theory to the time-based dimension, spillover theory to the strain-based dimension, and boundary theory to the behavior-based dimension. From this theorizing, we highlight and meta-analytically test the relative importance of specific variables from the WFC nomological network that are theoretically connected to each dimension: time and family demands for the time-based dimension, work role ambiguity for the strain-based dimension, and family-supportive supervisor behaviors and nonwork support for the behavior-based dimension. Reviewing and drawing from bandwidth-fidelity theory, we also question whether composite-based WFC approaches are more appropriate for broad constructs (i.e., job satisfaction and life satisfaction). The results of our meta-analytic relative importance analyses generally support a dimension-based approach and overall follow the pattern of results expected from our dimension-level theorizing, even when broad constructs are considered. Theoretical, future research, and practical implications are discussed. (PsycInfo Database Record (c) 2024 APA, all rights reserved).