Screening methods for the detection of metallo-beta-lactamase (MBLs)- producing isolates have been proposed. To investigate those methods, molecular techniques are needed as confirmatory methods for detection of MBLs genes. Simplex polymerase chain reaction (PCR), using individual sets of specific primers for MBLs genes has been studied. However, the technique is time-consuming and not cost-effective. Therefore, a better PCR technique is needed. The primary aim of this study was to develop a multiplex PCR for the detection of MBLs genes and to determine the prevalence of MBLs-producing ceftazidime (CAZ)-resistant gram-negative bacilli (GNB). The secondary aim was to take up Nuttiya Intachang’s question (in a previous study) on the comparison of two screening methods, double disk synergy test (DDST) and combination disk diffusion method (CBT). Two genetic methods were used, multiplex PCR (developed in this study) and dot blot hybridization. In the first section, the condition of multiplex PCR, using newly designed IMP, VIM, and L-1 primers in the same reaction, was optimized. In the second section, the two genetic methods were used to determine the MBLs genes in 100 CAZ and IMP-resistant GNB, and 541 CAZ-resistant GNB. These two methods were also used as confirmatory methods for analysis of sensitivity, specificity, and accuracy of two screening methods, DDST and CBT, as previously investigated by Nuttiya Intachang. MBLs genes-carrying GNB distributed in Ramathibodi Hospital, were found to predominate in non-glucose-fermenting bacteria, especially S. maltophilia, Acinetobacter spp. and Pseudomonas spp. The prevalence of blaIMP was higher than that of blaVIM. All of S. maltophilia were found to carry blaL-1. One percent of blaIMP or blaVIM and 23% of blaL-1 were found in CAZ and IMP-resistant GNB. One point eighty-five percent of blaIMP, 0.37% of blaVIM and 6.28% of blaL-1 were found in CAZ-resistant GNB. The multiplex PCR can detect MBLs gene, blaIMP, blaVIM, and blaL-1 simultaneously. This technique detects MBLs genes rapidly and is cost-effective. The sensitivity, specificity, and accuracy of the CBT developed by Nuttiya Intachang and DDST developed by Arakawa, were not different. However, the DDST had some limitations. The CBT is simple, inexpensive and may be used for the detection of MBLs-producing isolates in the routine clinical microbiology laboratory.