Measurement of the performance of mass spectrometers is required to determine which model best meets the requirements of a process. For this study, the instruments under test were adjusted according to manufacturers’ recommendations and these settings were maintained throughout the tests. Only Faraday cup data were collected. Scan speed/damping combinations were set to give low noise, moderate scan speed, and no peak clipping. Pressure decay through a molecular leak and calibrated leaks generated partial pressures from 10−4to 10−11Torr in a chamber pumped by conductance‐limited turbomolecular pumps. Ten parameters were measured. Peak shape and peak top structure were analyzed by least‐squares programs. Baseline and peak top noise were determined. Short‐ and long‐term stability of peak height, half peak height, and peak position were measured. Sensitivities to eight gases were determined. Linearities within and between electrometer ranges were calculated. The upper and lower pressure limits of linear response were found. Interferences of gases varied unpredictably. The instruments tested showed wide variations in performance.