In nonlinear model identification, mathematical modellers need to find the best input variables by training and testing all the likely model input combinations. This is very time consuming since a complete model development cycle is needed for each input variable combination. In this study, the Gamma Test (GT) is explored for its suitability in reducing model development workload and providing input data guidance before actual models are developed. The nonlinear dynamic model tested is the generalized regression neural network (GRNN). It has been found that the overall performance of the Gamma Test is quite encouraging and the GT demonstrates its huge potential for efficient GRNN model development. The Gamma values are able to provide a good indication about the achievable accuracy for the GRNN models and this has a distinctive advantage over the traditional model selection approaches.