Stable isotopes and mass-balance trophic models (e.g., ECOPATH) are well-known and widely used approximations to describe food-web structure, but their consistency is not properly established. Here we analyze the food-web structure of a subtropical-temperate coastal lagoon using two approaches: stable isotopic techniques and mass-balance modelling, exploring also the correspondence between the outputs of both methods. We compared trophic positions (TPs) derived by these two approaches for 14 consumers in Laguna de Rocha (LR).TPs based on stable isotopes were taken from a recent study. ECOPATH trophic levels were estimated by a model presented here constructed based on field data for the period 2003-2006 and literature data. The model incorporated over 50 species in 27 trophic groups, including primary producers, invertebrate and vertebrate consumers. The origin and quality of data (pedigree routine) indicated that 68% of the information was locally bound, although several unknowns were detected. Birds and mammals represented the highest trophic levels (4.2 and 3.98, respectively). Network analysis estimated a size of the system (fluxes and biomasses) of 451 t wet weight km(-2) year(-1), while transfer efficiency, primary production/respiration and production/biomass ratios, and several ecological indexes characterized LR as an underdeveloped system. TPs derived from isotopic analysis were highly correlated with trophic levels estimated by ECOPATH according to a linear regression model through the origin (r(2) = 0.82, n = 14, p << 0.01). The slope of the linear model (0.88 +/- 0.019, estimate +/- SD) indicated that TPs derived from isotopic analyses were slightly higher (similar to 13.5% on average) than those derived from the mass-balance model. Current results support the overall consistency of the use of both stable isotopes and mass-balance modelling approaches as descriptors of this aquatic food-web. (C) 2010 Elsevier B.V. All rights reserved.