A graphics processing system includes a graphics processing pipeline including at least an initial processing stage and a further processing stage. Data for a scene at a first resolution and data for the scene at a second resolution are processed in the initial processing stage. After processing data for the scene at the first and second resolutions in the initial processing stage, gaze-tracking data relating to a current positioning of at least one eye of a user of a virtual reality user device is received. At least one sub-region of the scene is identified from the gaze-tracking data. Data for the scene at the first resolution and only data corresponding to the identified at least one sub-region of the scene at the second resolution are processed in the further processing stage. The scene is rendered by combining the data for the scene processed in the further processing stage.