Macrobenthic data from estuaries along the Pernambuco coast in north-eastern Brazil were analysed to evaluate the effect of sieve mesh size (1.0 mm × 0.5 mm) and sampling depth (0–10 cm × 0–20 cm) on the description of infaunal communities, in an attempt to discuss standardized sampling procedures for different ecological studies objectives in these ecosystems. In general, the difference in sieve retention was less evident for biomass but was important for abundance: the 1.0 mm sieve retained only 27% of total individuals but 77% of total biomass. Regarding sampling depth, the 0–10 cm layer contained most individuals (94%) but contributed just 64% of the overall biomass. Although no strong differences in community structure were observed at most sites with the use of different sieves, the correlations among community dissimilarity using different meshes with environmental parameters (organic matter, total-N and microphytobenthos) indicated that the use of the 0.5 mm sieve will allow a better evaluation of the status of these estuaries. The results also highlight the importance of taking the vertical distribution of tropical macrofauna into account for quantitative estimates: for taxa composition and abundance the top layer is clearly essential, whereas for biomass the deeper layers should also be considered.