Background: Efforts to elucidate subtypes within depression have yet to establish a consensus. This study aims to rigorously compare different subtyping approaches in the same subject space, to quantitatively test agreement across subtyping approaches and determine if the different approaches are sensitive to different sources of heterogeneity in depression. Methods: We will implement six different data-driven subtyping methods developed in prior work using the same UK Biobank participants (N=3,724 depressed cases and N=2,631 healthy controls). The six approaches include two symptom-based, two structural neuroimaging-based and two functional neuroimaging-based techniques. The resulting subtypes will be compared based on subject assignment, stability, and sensitivity to subgroup differences in categories of demographics, general health, clinical characteristics, neuroimaging, trauma, cognitive, genetics, inflammation markers. Proposed Analyses: We will assess the subject agreement of the subtype solutions using Adjusted Rand Index, the stability of the subtype solutions using bootstrapping, and evaluate domain (clinical, biological, neuroimaging) sensitivity differences between subtype approaches by performing general linear models on phenotypes unseen by previous analysis comparing subtypes within each approach and determining statistical difference in variance explained across subtype approaches (using permutation testing).