This VE-822 ATM/ATR inhibitor method is used worldwide in monitoring programmes and can be considered to be the ‘gold standard’. Replica plating on the other hand is a more feasible and costeffective method to quantify resistance within animal populations since multiple isolates can be tested simultaneously for their resistance using agar plates containing an antimicrobial at a breakpoint concentration. In order to determine resistance on herd-level a multi-level approach is needed. One needs to know how many animals within the herd have to be sampled, and how many isolates per faecal sample have to be tested to get a representative resistance level. For this purpose the variation in resistance among isolates within a faecal sample and the variation among faecal samples within a herd have to be investigated and taken into account. Although several studies on prevalence of antimicrobial resistance within herds have been performed, only a few studies addressed the variation in resistance of isolates within and among faecal samples and used this information to investigate the effect of different sampling strategies on the precision of estimated resistance levels. In a study among dairy cows the variance in resistance to 12 antimicrobials was mainly attributable to variation among isolates. Based on this information four different sampling strategies were simulated and it was suggested that testing 3–4 isolates per cow was the best strategy to determine the prevalence of antimicrobial resistance at herd-level. Among finisher pigs the largest component of variance in resistance to tetracycline and gentamicin existed between pigs within the same pen. Bootstrap analysis based on these data revealed that for tetracycline at least 5 isolates per faecal sample need to be tested, whereas for gentamicin testing more than 10 isolates per faecal sample would result in precise estimates of resistance at herd-level. Among broiler chickens the variation of resistance among isolates was low meaning that focus should be on the number of animals sampled within a flock rather than on the number of isolates tested within one animal. These studies show that the composition of the variance influences the choice for a sampling strategy in order to measure antimicrobial resistance at herd-level. Due to species-specific factors that might play a role in the development and spread of antimicrobial resistant isolates within a herd conclusions from these studies may not be representative for the situation within veal calf herds. The aim of this study was to determine a sampling strategy based on the variation in proportions of resistant isolates within and among faecal samples from veal calves in order to estimate the prevalence of resistance on herd-level. For this purpose we quantified and compared the proportions of resistant commensal E. coli isolated from veal calves, using two different test methods and two different sample types. In search for a sampling strategy to estimate resistance levels within a veal calf herd we investigated the reliability of the replica plating test method by comparing the obtained results to broth microdilution as a reference. With replica plating isolates were tested for their resistance using breakpoint concentrations one or two twofold dilution steps higher compared to EUCAST epidemiological cut-off values. Nevertheless, this method has proven to provide results comparable to those obtained with broth microdilution for amoxicillin, tetracycline and tmp/s. For ciprofloxacin a significantly lower odds of resistant isolates by replica plating was found.