Effects of surface-water diversions on habitat availability for native stream fauna (fish, shrimp, and snails) are described for 21 streams in northeast Maui, Hawaii. Five streams (Waikamoi, Honomanu, Wailuanui, Kopiliula, and Hanawi Streams) were chosen as representative streams for intensive study. On each of the five streams, three representative reaches were selected: (1) immediately upstream of major surface-water diversions, (2) midway to the coast, and (3) near the coast. This study focused on five amphidromous native aquatic species (alamoo, nopili, nakea, opae, and hihiwai) that are abundant in the study area. The Physical Habitat Simulation (PHABSIM) System, which incorporates hydrology, stream morphology and microhabitat preferences to explore relations between streamflow and habitat availability, was used to simulate habitat/discharge relations for various species and life stages, and to provide quantitative habitat comparisons at different streamflows of interest. Hydrologic data, collected over a range of low-flow discharges, were used to calibrate hydraulic models of selected transects across the streams. The models were then used to predict water depth and velocity (expressed as a Froude number) over a range of discharges up to estimates of natural median streamflow. The biological importance of the stream hydraulic attributes was then assessed with the statistically derived suitability criteria for each native species and life stage that were developed as part of this study to produce a relation between discharge and habitat availability. The final output was expressed as a weighted habitat area of streambed for a representative stream reach. PHABSIM model results are presented to show the area of estimated usable bed habitat over a range of streamflows relative to natural conditions. In general, the models show a continuous decrease in habitat for all modeled species as streamflow is decreased from natural conditions. The PHABSIM modeling results from the intensively studied streams were normalized to develop relations between the relative amount of diversion from a stream and the resulting relative change in habitat in the stream. These relations can be used to estimate changes in habitat for diverted streams in the study area that were not intensively studied. The relations indicate that the addition of even a small amount of water to a dry stream has a significant effect on the amount of habitat available. Equations relating stream base-flow changes to habitat changes can be used to provide an estimate of the relative habitat change in the study area streams for which estimates of diverted and natural median base flow have been determined but for which detailed habitat models were not developed. Stream water temperatures, which could have an effect on stream ecology and taro cultivation, were measured in five streams in the study area. In general, the stream temperatures measured at any of the monitoring sites were not elevated enough, based on currently available information, to adversely effect the growth or mortality of native aquatic macrofauna or to cause wetland taro to be susceptible to fungi and associated rotting diseases.