This paper presents a study on attitude estimation during the Plasma Brake Experiment (PBE) onboard a small satellite. The PBE demands that the satellite be spun at a very high angular velocity, up to 2001deg/g, to deploy the tether using centrifugal force. The spin controller, based on purely magnetic actuation, and the PBE demands accurate attitude estimation for the successful execution of the experiment. The biases are important to be estimated onboard small satellites due to the closely integrated systems and relatively higher interference experienced by the sensors. However, bias estimation is even more important for PBE due to the presence of a high voltage unit, onboard the satellite, that is used to charge the tether and can be the source of interference. The attitude and the biases, when estimated simultaneously, results in an augmented state vector that poses a challenge to the proper tuning of process noise. The adaptation of process noise covariance has, therefore, been studied and analysed for the challenging PBE. It has been observed that adapting the process noise covariance improves the estimation accuracy during the spin -up phase. Therefore, it is very important to use adaptive process noise covariance estimation. (C) 2016 IAA. Published by Elsevier Ltd. All rights reserved.