Sampling was seasonally performed in the Patos Lagoon estuary (Rio Grande, RS, Southern Brazil) to estimate zooplankton biomass and production comparing values obtained using an enzymatic (chitobiase) method and the traditional mathematical models based on growth. Comparison of data obtained from zooplankton samples collected with 90 and 200-μm mesh nets showed that net selectivity influences the estimation of zooplankton biomass and production. Furthermore, it showed differential results for dominance of taxa and proportions of developmental stages in samples. Differences among samples collected at the different sites in the same season were observed using either the mathematical models or the enzymatic method. The two different approaches were also able to detect the seasonal variation in production. In a broad view, data obtained using the chitobiase method showed a similar pattern of zooplankton production compared to those obtained with the traditional mathematical models based on growth. However, values estimated using the enzymatic method were systematically higher than those obtained with the mathematical models. Maximum total production values were estimated as 12.5, 9.2 and 7.9mgCm−3day−1 for the “chitobiase method”, “Huntley model”, and “Hirst model”, respectively. Considering all sampling sites and seasons, the magnitude of this difference corresponded to 1.95 and 2.49mgCm−3day−1 for the “Huntley model” and the “Hirst model”, respectively. These findings indicate the reliability of the enzymatic method in estimating crustacean production also in estuarine environments of changing salinity, as previously demonstrated for marine waters. In addition, the use of this method is more practical and comparatively less time-consuming and cheaper than the use of the mathematical models based on growth.