Except for two or three target parameters， all the remaining parameters（hereinafter referred to as background parameters） are usually assumed to be accurate in parameter identification. However， both models and operating conditions are highly random and uncertain in nature， typical or experiential values have to be used for the background parameters. It is also difficult to assign reasonable value-intervals for some background parameters. Errors of the background parameters will affect the identification errors of the target parameters， or even invalidate the identification. Based on stability mechanisms， the influence of background parameter uncertainty on parameter identification by using the difference index and its sensitivity analysis are investigated. Simulation results show that background parameter error may lead to serious error for identification or even no solution. This reveals the potentially very high risks for parameter identification.