Parameter estimation in power systems is often constrained by the number and location of measurement devices in engineering applications, leading to some parameters with high estimation accuracy, while other parameters exhibit large estimation error. Motivated by this phenomenon, the theory and analysis approach of engineering testability is proposed to effectively identify parameters with application value. The relevant concepts of engineering testability are defined first. Under different measurement conditions, engineering testability can be classified into three kinds: intrinsic testability, measuring testability, and disturbing testability. Then, an analysis approach of testability based on Monte Carlo method is expounded. By comparing error index values with the prescribed threshold, it can efficiently identify whether the parameters to be estimated are clear or uncertain, and the testable degree of the system can be calculated. Simulations on parameter estimation of grounding grid and distribution network illustrate that the utilization of testability can identify clear and uncertain parameters in the system. Moreover, estimation values of clear parameters and confidence intervals of some disturbing uncertain parameters can provide a beneficial reference for engineering applications.