Sir, Currently, a variety of qnrA, B and S genes are commonly isolated from clinically important Enterobacteriaceae. Moreover, qnr genes have been detected in non-typhi Salmonella enterica serotypes in Europe, the United States, Africa, Australia and Asia. In a recent study, the presence of qnrA1 in the Netherlands was first detected in a multidrug-resistant Enterobacter cloacae carrying a conjugative R plasmid originating from a large outbreak in the University Medical Centre Utrecht (UMCU). In The Netherlands, data on the occurrence of qnr genes in other Enterobacteriaceae, such as S. enterica, are lacking. The aim of our study was to detect and characterize qnr genes in a selection of Dutch Salmonella isolates. The strain collection of the Central Institute for Animal Disease Control encompasses 15 011 Salmonella isolates collected from humans (n 1⁄4 8143), poultry (n 1⁄4 2570), pigs (n 1⁄4 992), cattle (n 1⁄4 518) and other sources (n 1⁄4 2788) in the period 1999–2006. Thirty-nine of these isolates showed a typical phenotype being low-level resistant to ciprofloxacin (MICs: 0.25–1 mg/L) but still susceptible to nalidixic acid (MICs: 8–16 mg/L). All isolates showing this typical phenotype were screened for the presence of qnrA, qnrB and qnrS genes by PCR using primers and conditions as previously described. Transconjugants of Escherichia coli J53 harbouring the plasmids pMG252 (qnrA1) or pMG298 (qnrB1) and S. enterica serotype Bovismorbificans strain AM12888 (qnrS1) were used as positive controls. In addition, all isolates were screened for the presence of aac(6 0)-1b-cr by PCR as described by Robicsek et al. using transconjugant 10–2 containing plasmid pHSH10-2 as a positive control. Gel electrophoresis revealed 3 qnrB-positive isolates and 31 qnrS-positive isolates. Five isolates were PCR negative for all qnr genes. On these isolates, a second PCR using degenerate qnrB primers as described by Cattoir et al. did not yield any specific products. No qnrA and aac(60)-1b-cr genes were detected in any of the isolates. Sequence analysis of the 34 PCR products with BLAST (http://www.ncbi.nlm.nih.gov/BLAST/) revealed two variants of the qnrB gene (qnrB2 and qnrB50) in two different serotypes and one variant of the qnrS gene (qnrS1) in six different serotypes, predominately Salmonella Corvallis (n 1⁄4 25) (Table 1). Except for one qnrB2-positive Salmonella Bredeney isolated from a Dutch broiler chicken, all qnr-positive isolates were of human origin. Additional sequence analysis of gyrA, gyrB, parC and parE PCR products revealed no mutations in the quinolone resistance-determining region (QRDR). This is the first report of the occurrence of qnrB (qnrB2 and qnrB50) and qnrS1 genes in The Netherlands, and also the first report of the occurrence of qnr genes in S. enterica strains isolated from humans as well as a poultry source from The Netherlands. Although human non-typhoidal S. enterica predominantly have an animal origin, the source of the qnr genes in these isolates is uncertain. The occurrence of qnr genes in Salmonella isolated from German poultry has been reported earlier by Kehrenberg et al. The presence of qnr genes in poultry is a reason for concern, because the use of fluoroquinolones in these animals may facilitate the horizontal transmission of qnr genes in the gastrointestinal tract and thus increase the reservoir. In this study, the qnrS1 variant was predominantly present in Salmonella Corvallis. qnrS1-positive Salmonella Corvallis isolates originating from Denmark and Thailand have recently been described. PFGE analysis of the 25 qnrS1-positive Dutch Salmonella Corvallis isolates revealed 9 different subtypes in 3 major clusters (data not shown). Except for one cluster (two isolates), the Dutch and Danish PFGE patterns show a high similarity, indicating common sources for the qnrS1-positive Salmonella Corvallis isolates. The high detection rate of qnr genes in the strain selection demonstrates that isolates harbouring the qnr gene without any mutations in the QRDR region can very well be recognized