Abstract The Diavik diamond mine includes four diamondiferous kimberlite pipes (A154N, A154S, A418, and A21) and minor kimberlite dikes that are mined by Rio Tinto and Dominion Diamond Corporation. Pipe morphologies from A154S, A154N, and A418 show similar circular near-surface expressions, with expansion of cross-sectional area and elongation at depths below 100 m above sea level, consistent with the dominantly steeply dipping to nearly vertical structures in the area (~050°). The internal geology of the pipes is highly variable; twenty-nine distinct domains form the basis for three-dimensional geologic models. The kimberlite deposits at Diavik are highly variable and suggest that emplacement into similarly shaped pipes within a single cluster can vary significantly, resulting from a single infilling sequence (e.g., A154S), multiple events separated by periods of relative volcanic quiescence (e.g., A154N), or changes in magma flux or hydrologic conditions (e.g., A418). Diavik pipe emplacements likely result from a shared five-stage emplacement continuum involving (1) exploitation of existing structures by early kimberlite magma intrusion, (2) initial pipe excavation, (3) pipe infilling, (4) sedimentation into craters, and (5) late magmatic intrusion. Indicator minerals from each kimberlite pipe contain varying proportions of the same mantle source. Differences in the relative abundances of mantle minerals are shown among different pyroclastic domains in a single pipe, suggesting emplacement via multiple magma pulses containing unique mantle populations, while units from different pipes are, in some cases, more similar and reflect cross-fertilization among adjacent pipes. Clinopyroxene thermobarometry indicates that the Diavik garnet lherzolite xenoliths were sampled from 100- to 200-km depths and suggests that magmas emplaced at A154N successively sampled deeper mantle through time. Structural analysis indicates that faulting occurred prior to kimberlite emplacement in varying stress regimes. Faults were subsequently reactivated or exploited by early-stage ascending kimberlite magmas, impacting both pipe elongation directions and excavation depths during emplacement. Exploration assessments on the basis of shallow (<150 m) drilling alone may underestimate the volume and/or diamond potential, particularly for isolated (i.e., unclustered) bodies.