Surface coal mining alters landscapes including creating waste-rock fills or dumps. In Appalachia USA, mines fill valleys with waste rock, constructing valley fills that affect water quality and aquatic ecology downstream. Total dissolved solids (TDS) in mine effluent are elevated from exposure of mineral surfaces to weathering. Understanding TDS variability requires understanding valley fill internal structure and its effect on hydrology, yet prior studies focused on point measurements or did not address patterns among fills. Here we investigated subsurface structure and hydrologic flowpaths in two dimensions within four valley fills using electrical resistivity imaging (ERI). We used artificial rainfall to investigate the location and transit time of preferential flowpaths through the fills. We corroborated our ERI interpretations using borehole logs, downhole video, and shallow soil excavation. ERI results indicated variability in substrate type and widespread presence of preferential flowpaths. We estimated an average preferential flowpath vertical length of 6.6 m, average transit time of water along the flowpath of 1.4 h, and average minimum water velocity of 5.1 m/h (0.14 cm/s). These rates are higher than typical for undisturbed lands, and resemble highly preferential flow in karst terrain. ERI successfully distinguished fills using conventional loose-dump construction from experimental controlled-material compacted-lift construction. Conventional fills exhibited finer particles that retain water at the surface, with larger rocks and larger voids at depth. Conventional fills had greater ranges of subsurface resistivity (i.e. substrate types) and greater interior accumulation of water during artificial rainfall, indicating more quick/deep preferential infiltration flowpaths. We show experimental construction significantly alters hydrologic response, which in combination with use of low-TDS waste rock, may affect downstream water quality relative to conventional loose-dump methods. Our soil boring and pits corroborated ERI interpretation, thus demonstrating ERI to be a robust non-invasive technique that provides reliable information on valley fill structure and hydrology.