Asphaltene deposition was investigated during laboratory-scale steam-assisted gravity drainage (SAGD) experiments to probe in situ upgrading of a heavy oil. Tests were conducted with and without the addition of non-condensable gases (carbon dioxide or n-butane) to the steam. The apparatus was a three-dimensional scaled physical model packed with crushed limestone saturated with 12.4 degrees API heavy-crude oil. Temperature, pressure, and production data, as well as the asphaltene content of the produced oil, were monitored continuously during the experiments. For small well separations, as the fraction of non-condensable gas in the steam increased, the steam condensation temperature and the steam-oil ratio decreased. As a result of lower temperature, the heavy oil was less mobile in the steam chamber relative to pure steam injection. Thus, the heating period was prolonged and the recovery, as well as the rate of oil recovery, decreased. Asphaltene content of the oil produced as a result of pure steam injection decreased initially showing deposition of asphaltene within the porous matrix of the model. As the steam injection continued, the asphaltene content of the produced oil increased but remained below the initial value. Thus, the produced oil indicated some in situ upgrading. As the carbon dioxide concentration in the steam increased, greater asphaltene deposition occurred; however, no significant change in asphaltene content was found when n-butane was added to the steam. Post-experimental analyses of the porous media for asphaltene content confirmed retention for the pure steam and steam with added CO 2 experiments. Numerical simulation of the asphaltene deposition process using a pure solid deposition model corroborated experimental findings and showed that deposition occurred mainly at the steam-chamber boundary.