The tetracyclines (TCs) are widely used in the treatment of several diseases of cattle and their residues may be present in milk. To control these residues it is necessary to have available inexpensive screening methods, user-friendly and capable of analysing a high number of samples. The purpose of this study was to design a bioassay of microbiological inhibition in microtiter plates with spores of Bacillus pumilus to detect TCs at concentrations corresponding to the Maximum Residue Limits (MRLs). Several complementary experiments were performed to design the bioassay. In the first study, we determined the concentration of spores that produce a change in the bioassay's relative absorbance in a short time period. Subsequently, we assessed the concentration of chloramphenicol required to decrease the detection limit (DL) of TCs at MRLs levels. Thereafter, specificity, DL and cross-specificity of the bioassay were estimated. The most appropriate microbiological inhibition assay had a B. pumilus concentration of 1·6 × 109 spores/ml, fortified with 2500 μg chloramphenicol/l (CAP) in Mueller Hinton culture medium using brilliant black and toluidine blue as redox indicator. This bioassay detected 117 μg chlortetracycline/l, 142 μg oxytetracycline/l and 105 μg tetracycline/l by means of a change in the indicator's colour in a period of 5 h. The method showed good specificity (97·9%) which decreased slightly (93·3%) in milk samples with high somatic cell counts (>250 000 cells/ml). Furthermore, other antimicrobials studied (except neomycin) must be present in milk at high concentrations (from >5 to >100 MRLs) to produce positive results in this assay, indicating a low cross specificity.