Fluid thrust vectoring technology is regarded as a key technology for the development of very low detectable vehicles because of its advantages such as fast response， light weight and good stealth performance. The current research on the performance of fluid thrust vectoring nozzles is mainly focused on univariate studies. Multivariate sensitivity analysis of fluid thrust vectoring nozzles is carried out to help clarify the key parameters affecting the nozzle vectoring performance， and guide the design of fluid thrust vectoring nozzles. The sensitivity analysis method based on non-intrusive polynomial chaos was carried out to investigate the global sensitivity analysis of seven design variables such as Mach number of outflow， jet position， jet angle， Nozzle Pressure Ratio （NPR） and correlation analysis for different targets of the two-dimensional Shock Vector Control （SVC） nozzle. The results show that vector angle and vector efficiency are significantly sensitive to jet position； thrust coefficient is more sensitive to NPR， outflow Mach number， secondary flow pressure ratio （the ratio of total secondary flow pressure to ambient pressure SPR_{t}）. The results also show that secondary flow ratio is more sensitive to NPR， and the contribution of ratio of total mainstream and secondary flow temperature to ambient temperature to the vector performance is expressed mainly in the secondary flow ratio. The increase of both vector angle and secondary flow ratio leads to thrust loss， and the secondary flow ratio has the greatest impact. The increase in outflow Mach number changes the actual pressure ratio of the nozzle and reduces the nozzle vectoring performance. When the position of the secondary flow is close to the throat， the separation before the jet extends to the throat， and the separation after the jet reattaches under the large NPR. The change of the nozzle aerodynamic profile results in the decline of the vector performance， and in serious cases， the vector angle reversal occurs. The SVC nozzle should be designed so that the jet is positioned close to the nozzle outlet and the oblique shock wave in the nozzle is adjusted to develop near the nozzle outlet for a given NPR and SPR_{t}， thus improves the vector performance without increasing the secondary flow ratio. A reasonable combination of parameters can mitigate the negative impact of increasing outflow Mach number on nozzle vector performance， and can improve nozzle vector performance without increasing the secondary flow ratio.