HVAC systems are composed of different noise sources and paths. The non-attenuated propagation of noise through the system has detrimental effects on acoustic comfort of people inside the premises. In order to mitigate the propagated noise, classic parallel baffle splitters are used to reduce the transmitted noise through acoustic coatings. Different methods have been developed to predict the insertion loss of those elements, however, if the input data is not well known these models can lead to deviated results. Additionally, the use of splitters in HVAC systems produces pressure drop which can damage the installed mechanical equipment if that is not well predicted. Currently, the models that estimate pressure drop relate dimensional features and design velocity. However, these models can show overestimated results. In this work an experimental rig was implemented to assess a splitter installed inside of a test duct. Measurements were performed to estimate insertion loss and pressure drop coefficient, following the guidelines exposed on the ISO 7235 standard. The results were compared with analytical methods. Finally, a numerical method analysis of the test rig was performed, showing the correlation between these results and the experimental data.