This thesis examines CubeSat attitude determination using the Earth’s magnetic field (EMF) vector aiding a low-cost IMU. CubeSats provide relatively cost-effective methods of performing scientific research in orbital environments. However, to adequately perform this research, knowledge of the CubeSat’s orientation in 3D space (attitude) is often required. To that end, the design of a reliable attitude determination (AD) system on-board a CubeSat is a critical aspect for many mission designers. As a primary goal of CubeSat design is to ensure science objectives are met while minimizing, cost, mass, and volume, this thesis investigates a minimal sensor approach to CubeSat AD. Specifically, an inertial based AD scheme reliant on the use of an inertial measurement unit (IMU) aided only by vector measurements of the Earth’s magnetic field (EMF) is developed. An extended Kalman filter (EKF) approach to recursively estimate the attitude on-orbit using an IMU and a three-axis magnetometer (TAM) is detailed. Additionally, we describe a test to assess the stochastic observability of the EKF developed. We present simulation results showcasing the performance of the AD system for multiple orbital inclinations and initial attitude errors. Moreover, we discuss conditions in which the EMF vector can and cannot be effectively utilized as the sole aiding measurement, and we evaluate the stochastic observability of the linearized discrete time system. We extend the AD system discussed here to two current University of Minnesota Small Satellite Research Lab CubeSat designs: IMPRESS and EXACT. We describe future work for the implementation of the AD system and potential improvements to the EKF design.