This paper summarizes a set of SAGD experiments conducted live under an X-ray scanner. These experiments were specifically designed for mapping noncondensable gas distribution and their movement in an active steam chamber during SAGD. Many researches over the past 3 decades have shown that noncondensable gases may have critical impacts on SAGD performance. Some may be positive and others may be negative, depending on reservoir and operating conditions. To better use the positives, avoid the negatives, and for better SAGD performance predictions, it is crucial to understand how these gases behave in a steam chamber. It is arguable that noncondensable gases tend to accumulate at the steam front where steam condenses. However, this assertion has only been supported by numerical simulations. Field observation data have been too sparse. Meaningful tracking of gas production is not a normal practice in the field. The first experiment was conducted in an aluminum vessel packed with 4 darcy sands at 1.0 MPa. The second experiment was conducted in a scalable system consisting of a titanium pressure vessel and a PEEK cell, allowing the SAGD experiment to run at 2.1 MPa. Both experiments used bitumen fully saturated with methane at reservoir conditions and were run live under the X-ray scanner. X-ray images were taken at given time intervals. Temperature profiles were obtained directly from thermocouples. Density profiles were computed from the X-ray images. Methane in the free gas phase were calculated and mapped. After each experiment, samples from the opened cell were also tested for additional observation and confirmation. These experiments confirmed the assertion that noncondensable gas tends to concentrate along the steam front. It was also demonstrated that the steam temperature zone does not coincide with the oil-depleted zone, indicating that in a SAGD reservoir with non-trivial presence of noncondensable gases, temperature measurements at observation wells alone would not reflect the boundary of the steam chamber. The more representative measure of a steam chamber should be the mapping of the oil-depleted zone. A more comprehensive monitoring of gas production plus 4D seismic would be needed to determine the oil-depleted zone in the field operation.