Loss Of Control In-flight (LOC-I) is currently the largest cause of accidents for the world wide commercial jet transport fleet. To decrease the occurrence of this type of accident, procedures have been modified and pilot training has been increased. However, current training devices lack the appropriate flight models and motion capabilities to be fully representative of these accidents. With the SUPRA generic transport aircraft model and the Desdemona simulator it is possible to provide a representative flight model and to simulate the positive g-loads experienced during upsets and upset recoveries. To better understand the pilot response in these manoeuvres a descriptive pilot model was developed in this thesis. This provides insight into which indications are used by the pilot and how they are interpreted. Additionally the effect of three simulated motion conditions was explored; no motion, extended 'hexapod-like' motion and centrifuge motion. Ten civil transport pilots from several different airlines with varying levels of experience flew two types of stall scenarios in the Desdemona simulator. Both scenarios were repeated 12 times with variations in cockpit indications, control loading and motion condition. Based on the experiment data and pilot feedback a basic descriptive pilot model was developed. Using an iterative fitting process the pilot model response was matched to the experiment data. The resulting pilot model provides a reasonable fit for the average pilot response in the no motion condition. The variation between experiment runs can partially be reproduced by varying the pilot model control gains. No accurate representation could be determined for the effect of the speed tape and the pilot model does not fully reproduce the typical discrete control input behaviour. The experiment also showed a significant effect of the extended 'hexapod-like' motion condition on pilot control behaviour in the unloading phase of the stall recovery. A significant effect of the centrifuge motion condition was found in both the unloading and loading phase of the stall recovery. These effects could not be reproduced by classical negative motion feedback in the pilot model, but were best represented by a small decrease in the pilot model open loop control gain.