3D systems due to its complicated electronical, mechanical and vision accessories have enormous degree of complexity both in design and evaluation. Navigation system usually plays an important role in most 3D systems. Therefore, having an efficient navigation system and evaluation of the system has great deal of importance. 3D systems have a lot of parameters, inputs and outputs thus; evaluation process requires a complete test bench to collect and analyse maximum possible efficient parameters and variables. This paper will propose a test bench based on a tracker, navigation device status, and some biological and motion feedback from end-user to analyse a 3D system and associated navigation system performance. An original test procedure will be provided for evaluating navigation in 3D scenes by focusing on experimental aspects such as metric variable definition for measurement, case selection criteria and test-bed design. The novelty in our approach is the use of information given by the brain to understand the human perception in VE and to improve navigation in 3D scenes and interaction with virtual reality.