This thesis describes the analysis, design and construction of a two- degree of freedom (pitch and roll) fixed base spacecraft simulator and presents measurements of the human pilot's ability to perform attitude acquisition tasks in an inertial system, while using star constellations as a visual reference. Three subjects, with no prior flying experience, attempted to correct an initial attitude error in either minimum time or with minimum control effort (minimum fuel), using maximum angular acceleration levels of 1°/sec. and 4°/sec.. The control system utilized on-off torque thrusters to generate constant angular acceleration levels. The fuel used by the torque thrusters was a measure of the pilot's control effort. The subjects were able to "fly" the inertial system after a short familiarization period. The subjects attempting to perform the task in mini mum time tended to: 1. Keep the product of (acquisition time)(fuel required) a constant for the two acceleration levels . 2. Approximate a true optimum bang-bang control (which is minimum time) more nearly at the lower acceleration capability than with the higher one. In the minimum control effort task, the control system tended to be used the same percentage of the total time for both acceleration levels. When performing the minimum time and minimum control effort tasks, 66% and 50%, respectively, of the acquisition time was spent within ± 3.5° of the desired attitude although the initial attitude error magnitude was = 30° and = 23°. The subjects encountered more difficulty detecting angular rate and position errors in the roll axis than in the pitch axis.