Current melting of glaciers has increased interest in understanding glacier hydrology and the interplay between subglacial meltwater, the glacier bed, and ice behavior. However, little is known about the sedimentology of subglacial meltwater deposits, and no sediment type or sequence has been identified as being unique to subglacial settings. Here, we analyze a repetitive, coarsening-upward, sorted-sediment sequence in a thick, diamicton section that we interpret to be seasonal, subglacial meltwater deposits. Over 20 m of diamicton that we interpret to be subglacial traction till is exposed in a rock-cored drumlin at Dösebacka, Sweden, and the till contains multiple interbeds, 25–75 cm thick, of sorted sediment, each consisting of a coarsening-upward sequence of laminated clay, graded laminae of silt, and sand with pebbles. Based on sediment logs, grain-size, thin-section analysis, structural analysis, and field sedimentology, we investigate the genesis of these interbeds and conclude the interbeds to represent subglacial pond deposits. The coarsening-upward sequence represents cessation of till deposition due to hydrostatic bed separation followed by (1) suspension settling of clay in a subglacial blister/pond with drop-grains from an ice roof, (2) suspension settling of pulses of silt and very fine sand in still water coupled with occasional turbidites, (3) traction sedimentation (and loading) of sand and gravel from flowing water, and (4) renewed till deposition following reattachment. The interbeds also reveal deformation that is oriented parallel to the ice-flow direction and which we attribute to deformation during ice-bed reattachment.This sediment sequence parallels in character the seasonal behavior of supraglacial meltwater development on Greenland and moulin drainage, and we argue that the interbeds represent pond deposits during seasonal meltwater drainage. Drainage formed a subglacial blister/pond when the ice reached floatation point, and clay and, later, silt were deposited. With time, drainage became better integrated, and sand and pebbles were deposited by flowing water until reattachment when till deposition resumed.Additionally, based on our interpretation and the attitude of the beds in the drumlin, which parallel the drumlin slope, we argue the drumlin to have been formed by accretion rather than erosion. The anomalous thickness of the till and the numerous interbeds also imply a relatively rapid sedimentation rate, indicating the drumlin formed quickly (10's of years). This stoss-side environment allowed for preservation of the unique interbeds.