Depending on the configuration of the wind tunnel, there are generally two types of methods to use smoke for flow visualisation. The first is a handheld/traverse mounted smoke nozzle. This generates a single stream of smoke and can be positioned so as to focus on an area of interest on the tunnel model. The second method is having an installed rake of nozzles upstream from the test section, which each provide a smoke stream. A diagram of a rake installed upstream of the test section is shown below. The rake is shown from floor to ceiling in this case, with a single "sheet" of smoke traces travelling down to the test section.

The test article is then immersed in the flow and the behaviour of the stream traces is observed. For 2D testing, the wing section should span the test section from wall to wall to best visualise stall. For tests of aircraft, the models can be inserted the same way, or with the wingtips oriented towards the floor and ceiling of the test section. This decision is driven by what the test wishes to show, as well as trial and error with model positioning. It is important to remember that if the model is inserted through the wall, the model should be aligned with the smoke sheet.

The test procedure is similar to the other tests discussed, where a pre-determined range of AoA or AoS is swept through, with observations made at each test point. Images of tests conducted in the smoke tunnel are shown below.