Pelagic fish assessments often combine large amounts of acoustic-based fish density data and limited midwater trawl information to estimate species-specific biomass density. We compared the accuracy of five apportionment methods for estimating pelagic fish biomass density using simulated communities with known fish numbers that mimic Lakes Superior, Michigan, and Ontario, representing a range of fish community complexities. Across all apportionment methods, the error in the estimated biomass generally declined with increasing effort, but methods that accounted for community composition changes with water column depth performed best. Correlations between trawl catch and the true species composition were highest when more fish were caught, highlighting the benefits of targeted trawling in locations of high fish density. Pelagic fish surveys should incorporate geographic and water column depth stratification in the survey design, use apportionment methods that account for species-specific depth differences, target midwater trawling effort in areas of high fish density, and include at least 15 midwater trawls. With relatively basic biological information, simulations of fish communities and sampling programs can optimize effort allocation and reduce error in biomass estimates.