In this report, proposals to improve the analysis and reporting of data in the European Union Summary Reports on antimicrobial resistance in animals and food collected from the Member States have been elaborated. A logistic regression modelling approach is recommended to assess significance of temporal trends in occurrence of resistance. Temporal trends in full scale minimum inhibitory concentration distributions should be also explored to increase the early detection of decreased susceptibility. The recommendations for the reporting of quantitative resistance data obtained through standardised dilution methods and the interpretation of resistance using epidemiological cut-off values are reinforced. A further harmonisation of national monitoring designs would facilitate data interpretation at European Union level, where weighted indicators of resistance should be designed, accounting for prevalence of bacteria, occurrence of resistance and monitoring design at national level. It is also proposed to complement the harmonised panel of antimicrobials used for susceptibility testing and to emphasize the monitoring of resistance in indicator bacteria. It is considered essential that resistance data should be reported at isolate level in order to address the multi-resistance phenomenon. For this purpose, a definition and an analysis approach of the multi-resistance as well as a list of important co-resistance patterns have been put forward. The reporting of isolate-based data should facilitate obtaining more detailed data on the animal population/food category sampled, the number and the origin of samples, the bacterial subtypes and the antimicrobial concentration ranges tested. This would enable more in depth analyses to be conducted as regards the identification of factors associated with resistance and the geographical diffusion of the multi-resistance. Finally, recommendations are given on the comparison between new emerging resistance types in healthy humans and animals/food.