Pre-stimulus brain activity is thought to modulate visual perception. However, the underlying processes are strongly debated. Moreover, the role of pre-stimulus activity beyond tasks with single, simple stimuli is largely unknown. Here, we analyzed electroencephalography recordings (EEG) in the sequential metacontrast paradigm (SQM). In the SQM, a sequence of vertical lines is presented, creating a percept of two expanding streams of lines. When one line has an horizontal vernier offset, this offset is attributed to all the lines following in the stream and participants can easily discriminate it. If a second line in the stream has also a vernier offset but in the opposite direction, the offsets integrate and the relative dominance of the two features determines participants’ responses. We investigated whether responses are determined by pre-stimulus activity. We extracted pre-stimulus power and phase at different frequencies (4-40 Hz). First, we did not find that pre-stimulus activity determined the percept. Second, using a linear classifier we were able to decode the relative dominance between two offsets from post-stimulus evoked activity. Third, when we related pre-stimulus activity to the results of the classifier, we observed pre-stimulus effects. Our findings suggest that single-trial decoding could help to unravel pre-stimulus effects that standard analysis fails to detect.