Abstract Objectives To evaluate the performances of the QMAC-dRAST GN (Gram-negative) kit for rapid antimicrobial sensitivity testing (AST) and two other methods, directly on positive blood-culture broth (PBCB), by comparison with a reference method: the MicroScan method based on broth microdilution on colonies isolated on PBCB subculture. Methods In total, 156 samples were collected prospectively from blood cultures positive for a Gram-negative rod. Each sample was tested with four AST techniques: (i) the QMAC dRAST GN kit, (ii) the disc diffusion (DD) method, (iii) the MicroScan method applied directly to PBCB; and (iv) MicroScan with isolates from PBCB subculture, as a reference. Results For 124 PBCB containing Enterobacterales, overall essential agreement (EA) and categorical agreement (CA) between the QMAC-dRAST on PBCB and the reference reached 95.7% and 93.5%, respectively. There were 3.0% very major errors (VME), 4.0% major errors (ME) and 2.8% minor errors (mE). A comparison of MicroScan on PBCB and the reference yielded 98.8% EA, 98.5% CA, and rates of 0.6% VME, 0.9% ME and 0.7% mE. The DD method on PBCB gave a CA of 95.8% and rates of 1.7% for VME, 2.0% for ME and 1.9% for mE. Results were obtained more rapidly for QMAC-dRAST (median of 6 h 37 min versus 18 h for the MicroScan and DD methods on PBCB). Conclusions The QMAC-dRAST system provided rapid results well correlated with the reference method on PBCB containing Enterobacterales. Given the shorter time-to-results, the QMAC-dRAST system constitutes a fast and reliable alternative to conventional AST methods.