The paper presents experimental results on the impact of top water and gas caps on SAGD performance. The effect of the top thief zone on oil drainage rates and potential oil and steam loss into the top zone were measured. The study involved the use of a large scale high-pressure/high temperature experimental facility for injecting steam into an oil sand pack and measuring oil drainage rates and development of temperature ahead of the steam chamber. Numerical modelling was conducted to predict field scale performance using the CMG's STARS simulator. An elemental experimental approach was used in the study to simulate a generic reservoir in the Athabasca region with a pay zone thickness of 50 in and an overlying thief zone thickness of 8 in. In this approach, a reservoir element was selected close to the oil/top thief zone interface. The element was located ahead of an advancing steam front. In order to set the initial conditions of the laboratory element to be similar to those in the field, field scale numerical simulation was conducted to determine the temperature distribution in the element. The field scale temperature profile was established in the laboratory elemental model to represent the element's initial temperature before the start of steam injection during the experiments. The paper discusses the results from the study and highlights the potential implications of the top thief zone on SAGD applications. In addition, differences between gas cap and top water thief zones on impacting the thermal and recovery efficiency of the SAGD process are demonstrated.