To investigate the combined effects of decreasing taxonomic resolution (i.e. species, family, phylum), the use of different mesh-size (1.0 mm and 0.5 mm) and the type of samplers used (van Veen vs. corers taken by divers) on the quality of data obtained, a comparative study was undertaken with the overall aim of identifying cost efficient methods for routinely monitoring the ecological change caused by Mediterranean fish farming. The results clearly showed that information loss was relatively low as data were aggregated at higher taxonomic levels, particularly up to the level of family or even order. It was also found that the extra information gained by sieving samples through a 0.5 mm sieve did not improve the ability to distinguish the potentially impacted sites from the control stations. Finally, it was found that a relatively large proportion of the available information concerning the community structure such as abundance, biomass or diversity is lost when sampling is carried out with corers. A cost/benefit ratio analysis for the two sampling and the two sieving methods showed minimal values for the van Veen samples (for both sieve fractions) at the family level, indicating that analysis at this level gives the best balance between precision of the results and decrease in taxonomic effort. However, if the time needed to sort the samples is included in the analysis, then samples taken with corers using a 0.5 mm sieve and identified to families seems like a good compromise between precision and cost.