Abstract: Critical flow Venturi nozzles utilize the choking characteristics of compressible gas flows to provide a stable mass flow rate and therefore serve as key components in gas-flow primary standards and high-accuracy industrial metering systems. This paper presents a comprehensive review of research on critical flow Venturi nozzles from the perspectives of discharge coefficient theory and model evolution, throat-flow mechanisms and metrological key issues, numerical simulation methods, experimental methods and traceable verification, as well as future development trends and application prospects. First, the development from ideal inviscid models to semi-empirical correlations and standardized designs that account for viscous effects, boundary layers and real-gas behavior is summarized, and the differences between various models and the main sources of uncertainty in the discharge coefficient are analyzed. Then, the core physical mechanisms that affect metering accuracy inside the nozzle are discussed in terms of geometric-parameter sensitivity, thermal boundary layers and condensation, shock-boundary-layer interaction and premature unchoking phenomenon (PUP), together with a brief overview of current national experimental capabilities and typical uncertainty levels. In terms of numerical simulation, the review focuses on the evolution from steady RANS (including density-based solvers and the SST k-ω model) toward transition models (γ-Re_θ), non-equilibrium condensation Euler-Euler two-phase flow models, real-gas equations of state (PR-EoS, GERG-2008), and conjugate heat transfer models. The application of high-fidelity unsteady methods (URANS/LES) in revealing shock-boundary-layer interaction and PUP is also discussed. Regarding experimental validation, the paper presents the combined validation strategy using multi-source diagnostic techniques such as high-frequency dynamic pressure sensing (≥100 kHz), temperature-sensitive paint (TSP), schlieren / interferometry imaging, together with primary pVTt (Pressure-Volume-Temperature-Time) facilities. The potential of virtual metrology and digital twins for nozzle calibration is also highlighted. Finally, future trends are outlined regarding innovative geometry and material designs, coordinated development of numerical simulation and experimental diagnostics, unified discharge coefficient models with intelligent optimization, and multi-physics coupling and standardization. In particular, the prospects of physics-informed neural networks (PINNs) for constructing data-physics hybrid models are pointed out. The review shows that, with the rapid development of microelectronics, high-fidelity numerical methods, advanced experimental diagnostics, artificial intelligence and precision manufacturing, considerable room remains for accurate modeling and engineering application of critical flow nozzle flowmeters under large diameter, high Reynolds number, humid air and other complex operating conditions.