Abstract: The supersonic-supersonic ejector is a compact and highly efficient device widely used in pressure recovery systems for chemical lasers and supersonic wind tunnels. Its flow field structure and pressurization process differ significantly from those of traditional subsonic-supersonic ejectors. To investigate the flow characteristics, parameter matching rules, and ejector performance, experimental and numerical simulation methods were employed to study constant-section and variable-section supersonic-supersonic annular ejectors (S-SAEs). Results indicate that the minimum starting pressure of the annular ejector is proportional to the contraction ratio of the mixing chamber, whereas the minimum pressure in the blind chamber shows no significant correlation with the contraction ratio. The S-SAE can operate effectively under high-load conditions; however, at low entrainment ratios, the static pressure ratio becomes excessively large, causing the secondary flow to enter the mixing chamber at subsonic speeds, thereby converting the ejector into a subsonic-supersonic ejection state. Under the same entrainment ratio, reducing the Mach number of the secondary flow enhances mixing but increases total pressure loss during deceleration, leading to reduced ejector performance. The results also show that ejector performance in the S-SAE is not directly determined by mixing characteristics. Based on the observation that the primary and secondary flows in the S-SAE maintain a stratified flow state before the shock train, a one-dimensional theoretical method for calculating the critical contraction ratio of the mixing chamber was derived under the assumption of non-mixing between the flows. Comparisons with numerical simulations reveal that theoretical values are slightly lower than actual contraction ratios, with deviations of less than 18.7%.