Résumé : The purpose of this study was evaluation of the VAPChip assay based on the “Rapid-Array-PCR-technology” which targets 13 respiratory pathogens and 24 β-lactam resistance genes directly on respiratory clinical specimens. The first step included analysis of 45 respiratory specimens in order to calibrate and determine the threshold for target genes. The second prospective step involved 85 respiratory samples from patients suspected of nosocomial pneumonia collected in two academic hospitals over an 8-month period. Results of the VAPChip assay were compared to routine methods. The first step showed a large proportion of positive signals for H. influenzae and/or S. pneumoniae. For identification, discrepancies were observed in seven samples. Thresholds were adapted and two probes were re-designed to create a new version of the cartridge. In the second phase, sensitivity and specificity of the VAPchip for bacterial identification were 72.9% and 99.1%, respectively. Seventy (82%) pathogens were correctly identified by both methods. Nine pathogens detected by the VAPChip were culture negative and 26 pathogens identified by culture were VAPChip negative. For resistance mechanisms, 11 probes were positive without identification of pathogens with an antimicrobial-susceptibility testing compatible by culture. However, the patient’s recent microbiological history was able to explain most of these positive signals. The VAPChip assay simultaneously detects different pathogens and resistance mechanisms directly from clinical samples. This system seems very promising but the extraction process needs to be automated for routine implementation. This kind of rapid point-of-care automated platform permitting a syndromic approach will be the future challenge in the management of infectious diseases.