Volume measurements are essential for a wide variety of lake and reservoir applications, with recent technological and computational advances greatly improving volume estimate accuracy. Triangulated irregular network bathymetric models, created with spatially referenced depth data, are now commonly available for computer processing. Large datasets can easily be handled in geographic information system software, but field data acquisition is the principal time investment; therefore, determining minimum data density requirements associated with accurate volume estimates has great practical and economic relevance. We used spatially referenced hydroacoustic depth data acquired along parallel transects to assess survey resolution required to obtain accurate volume measures on 5 northeastern Washington lakes and reservoirs. Volume estimates decreased in each study waterbody as sampling resolution decreased, exhibited by significant (P < 0.0001) monotonic decreasing functions. Our study showed that hydroacoustic surveys with 50 m transect spacing were 99.0–99.6% (95% confidence interval, 98.3–100.5%) accurate compared to modeled volume. Across all study sites, volume estimates based on 300 m spaced transect surveys averaged 4.2 and 9.6% lower, with and without near-shore transects, respectively, when compared to 50 m spaced transect surveys. The addition of a near-shore transect increased the accuracy of volume estimates at large transect spacings, especially when large near-shore changes in depth were missed. As expected, higher sampling resolution increased volume estimate accuracy; however, our results suggest that survey effort should focus on topographically complex areas, such as near-shore areas, when time and/or economic constraints limit the acquisition of data.