Converging results in different scientific fields (behavioural ecology, fisheriesbiology, acoustic tagging, fisheries acoustics, behavioural modelling) suggest theexistence of “micro-groups” inside fish schools. These would comprise a few (5–10) fishmaintaining contact during a period long enough to allow individuals to recognise eachother. It is hypothesised that they would prefer to share the space with familiar ratherthan anonymous conspecifics. To evaluate whether acoustic methods could be used torecognise “micro-structures” inside fish schools and help test the “micro-group” hypothesiswe analysed acoustic data from anchovy schools off Peru, and gadoids in the North Sea.Data collection used a multibeam sonar (Reson SeaBat 6012). In the Peruvian case study,the sonar was mounted set horizontally on a drifting research vessel and the internalstructure of the schools of anchovies was analysed, although individual fish could not bediscriminated. In the North Sea case study, the sonar was orientated vertically above ademersal trawl to allow observation of individual fish entering the trawl. Geostatisticalanalyses were used to evaluate the existence of small spatial structures in anchovyschools. In these schools, “micro-structures” with a scale as small as 0.5 m were observedacoustically. For the gadoids nearest neighbour distance (NDD) measurements were carriedout, suggesting that the fish aggregated in small groups (2 to 25 individuals, with anaverage of 3.7 fish per group) in the trawl catches. The perspectives and limitations ofthese results are discussed.