Quantitative real-time polymerase chain reaction (qPCR) can be a convenient alternative to the Most Probable Number (MPN) methods to count VTEC in milk. The number of VTEC is normally very low in milk; therefore with the aim of increasing the method sensitivity a qPCR protocol that relies on preliminary enrichment was developed. The growth pattern of six VTEC strains (serogroups O157 and O26) was studied using enrichment in Buffered Peptone Water (BPW) with or without acriflavine for 4-24h. Milk samples were inoculated with these strains over a five Log concentration range between 0.24-0.50 and 4.24-4.50 Log CFU/ml. DNA was extracted from the enriched samples in duplicate and each extract was analysed in duplicate by qPCR using pairs of primers specific for the serogroups O157 and O26. When samples were pre-enriched in BPW at 37°C for 8h, the relationship between threshold cycles (CT values) and VTEC Log numbers was linear over a five Log concentration range. The regression of PCR threshold cycle numbers on VTEC Log CFU/ml had a slope coefficient equal to -3.10 (R(2)=0.96) which is indicative of a 10-fold difference of the gene copy numbers between samples (with a 100 ± 10% PCR efficiency). The same 10-fold proportion used for inoculating the milk samples with VTEC was observed, therefore, also in the enriched samples at 8h. A comparison of the CT values of milk samples and controls revealed that the strains inoculated in milk grew with 3 Log increments in the 8h enrichment period. Regression lines that fitted the qPCR and MPN data revealed that the error of the qPCR estimates is lower than the error of the estimated MPN (r=0.982, R(2)=0.965 vs. r=0.967, R(2)=0.935). The growth rates of VTEC strains isolated from milk should be comparatively assessed before qPCR estimates based on the regression model are considered valid. Comparative assessment of the growth rates can be done using spectrophotometric measurements of standardized cultures of isolates and reference strains cultured in BPW at 37°C for 8h. The method developed for the serogroups O157 and O26 can be easily adapted to the other VTEC serogroups that are relevant for human health. The qPCR method is less laborious and faster than the standard MPN method and has been shown to be a good technique for quantifying VTEC in milk.