Three groups of four Friesian cows in mid-lactation were fed a compound feedingstuff contaminated with 2, 10 or 300 mg chlortetracycline/kg for 21 days, and were then fed an uncontaminated diet for seven days. A fourth group of four cows was fed an uncontaminated diet throughout the study. Daily pooled milk samples from each cow were analysed by high performance liquid chromatography (HPLC) with a detection limit of 50 μg chlortetracycline/litre. Chlortetracycline was deteded in only two milk samples taken from one of the animals fed feed containing 300 mg chlortetracycline/kg; and both contained less than the maximum residue limit (MRL) specified by the European Union (100 μg/litre). All the milk samples were also analysed by the Delvotest sp microbiological assay, which has a detection limit of 300 μg chlortetracycline/litre. During the treatment period, this method gave four presumptive false-positive results, because they were not confirmed by HPLC. Seleded daily pooled samples from each treatment group were also analysed by the semi-quantitative Charm II radioreceptor assay with a detection limit of 10 μg chlortetracycline/litre. Immunoreactive chlortetracycline was detected only in the animals fed feed containing 300 mg chlortetracycline/kg and several of the results exceeded the EU MRL during the treatment period. No significant treatment effects on animal performance were observed. However, there was a trend towards a higher milk fat concentration (P<0.09) and a lower milk protein concentration (P<0.07) with increasing concentration of chlortetracycline in the diet.