One of the main sources of uncertainties in controlling the attitude of a satellite is the time delays seen in sensor data. Although it is possible to process sensor data to correct the deficiencies caused by delays, it is more suitable to design the controller robust enough to handle uncertainties well. In this study, the attitude of a 3-degrees of freedom satellite model, incorporating uncertainties (both sensor data delay and actuator misplacements), is controlled using a suitably designed fuzzy variable structure controller (FVSC). The performance of the FVSC is evaluated and compared to that of other reference controllers [proportional-derivative (PD), linear-quadratic-Gaussian (LQG), and loop-shaping controllers (LSCs)]. The FVSC performs well in both nondelayed and delayed (T-d = 0.2 s and 0.4 s) cases, while the PD and LQG controllers provide good response only for nondelayed cases (i.e. T-d = 0 s). A robust LSC also performs well in both cases, but its root mean squared error is high compared to the FVSC in the delayed case.