Larval lampreys are filter feeders that live for several years burrowed in fine sediments in freshwater streams. Stream side channels and edges, where larval lampreys gather, are vulnerable to natural and human-caused dewatering. Water level reductions can strand and kill thousands of larval lampreys, in part because many remain burrowed until their habitats are exposed, at which point larvae must emerge and attempt to move over dewatered substrate to locate wetted habitat. Dewatering for restoration efforts or seasonal closures of irrigation canals can be done slowly to reduce lamprey strandings, but in some settings, mechanisms are lacking to control the dewatering rate. Phased dewatering, where water level is reduced in stages separated by periods of static water level, could provide options when dewatering rate cannot be tightly controlled. To guide this phased approach, information is needed on the movement capability of larval lampreys. We examined larval lamprey (Entosphenus tridentatus and Lampetra spp.) movement distance and rate over dewatered substrate at shoreline slopes of 1%, 5%, 10% and 20% in a laboratory setting and modelled results using gamma regression models. Model results suggest both movement distance and movement rate increased with increasing slope and increasing larval length. We used the models to predict minimum distances and rates that 90%, 75% and 50% of medium-sized (75 mm) lampreys would move over dewatered substrates on slopes of 1%–20%. The models predicted that 50% of larvae could move distances of ≥31 cm at rates of ≥0.7 mm/s on a 1% slope and distances of ≥502 cm at rates of ≥8.6 mm/s on a 20% slope. We present an example scenario of how information on larval movement capabilities and shoreline slope could guide phased dewatering events to limit impacts to lampreys.