Fishery resource density is a key indicator for assessing resource status and is commonly estimated using bottom trawl survey data. Because of the non-uniform distribution of marine organisms, bottom trawl data always contain extreme values (zeros or very large values), which will affect the accuracy and precision of the estimate of resource density. In order to decrease the effect of extreme values on the estimation of resource density, the △-distribution model was used when analyzing trawl survey data for in the Yellow River estuary between 2013 and 2014.The estimate of the mean resources density using the model-based method and that using the designed-based method was compared, and the effect of the quantity and distribution of survey data on the results for the △-distribution model method was also discussed. With respect to survey data for in August and October, the density estimated using the model-based method was lower than that using the design-based method, while the densities estimated using the model-based method were higher than those using the design-based method for other data sets. The quantity and distribution characteristics of survey data might be important factors affecting assessment results using the △-distribution model. Variance estimated using the model-based method was much less than that estimated using the design-based method for all data for the two species in different months. Zero and maximum-value data had different effects on the estimation of resource density using the △-distribution model method; the model was robust when dealing with survey data with high stability and precision.