When performing fast LC with 50mm narrow-bore columns packed with small particles, the LC instrumentation can give rise to non-negligible band broadening. In the present study, the loss in chromatographic efficiency attributed to nine different mass spectrometers of various brands, ionization source geometries and types of analyzers was assessed. In their standard configurations, the extra-column variance of these UHPLC-MS systems was estimated to vary from 20 to >100μL(2). However, it was demonstrated that these differences arise exclusively from the chromatographic system (i.e., injector, tubing, valves, heater) and from the tubing employed to interface the UHPLC instrument with the MS device. By minimizing the tubing used for each UHPLC system, the extra-column variance was reduced to approximately 17-19μL(2) at 600μL/min, for all types of configurations. To achieve optimal chromatographic performance, it is therefore of prime importance to optimize the UHPLC configuration prior to conducting MS. The tubing located between the UHPLC system and the ionization source entrance was found to be particularly critical, as it contributes to band broadening even in the gradient mode. Using an optimized UHPLC-MS configuration, the loss in efficiency with a 50×2.1mm I.D. column was negligible for k>7. However, the efficiency loss with 1mm I.D. columns remained non-negligible for all current instrumentation, even for solutes with a value of k>20. Indeed, for a mixture of isobaric substrates and metabolites analyzed in gradient mode, the peak widths decreased by approximately 50% between a standard and optimized UHPLC-MS configuration, considering a 50×2.1mm, 1.7μm column. The peak broadening was changed by 230% on a 50×1mm, 1.7μm stationary phase, for the same system configurations.