Currently the most common imagery task used in Brain-Computer Interfaces (BCIs) is motor imagery, asking a user to imagine moving a part of the body. This study investigates the possibility to build BCIs based on another kind of mental imagery, namely "visual imagery". We study to what extent can we distinguish alternative mental processes of observing visual stimuli and imagining it to obtain EEG-based BCIs. Per trial, we instructed each of 26 users who participated in the study to observe a visual cue of one of two predefined images (a flower or a hammer) and then imagine the same cue, followed by rest. We investigated if we can differentiate between the different subtrial types from the EEG alone, as well as detect which image was shown in the trial. We obtained the following classifier performances: (i) visual imagery vs. visual observation task (71% of classification accuracy), (ii) visual observation task towards different visual stimuli (classifying one observation cue versus another observation cue with an accuracy of 61%) and (iii) resting vs. observation/imagery (77% of accuracy between imagery task versus resting state, and the accuracy of 75% between observation task versus resting state). Our results show that the presence of visual imagery and specifically related alpha power changes are useful to broaden the range of BCI control strategies. A brain-computer interface (BCI) typically translates the electrophysiological brain signals into an output that reflects the user's intent or mental activity. BCIs can be beneficial to people with severe motor disabilities 1 and may also be used beyond the scope of medical applications 2,3. BCI is not a perfectly accurate technology as it suffers from numerous issues, namely it is known to be a difficult task to reliably discriminate brain signal patterns from every subject. There are different approaches to improve the performance of BCIs. Most studies focus on signal processing and classification aspects. However, BCI performance can also be improved by optimizing the user's control strategies and therefore by identifying new and efficient mental tasks to achieve reliable control 4. Nowadays, the most common imagery task used in BCI is motor imagery, asking a user to imagine moving a specific part of the body such as a hand or a foot 5. It is used to control assistive technologies 6 , or even computer games 7. However, motor imagery based BCIs are seldom used outside laboratories due to their lack of reliability. The following main findings can be highlighted. First, around 20% of the BCI users are not able to obtain effective control. This phenomenon is being sometimes referred to as "BCI illiteracy" 8. Second, for even the remaining 80% of the people, motor imagery might not be the best mental task to control a BCI (not very intuitive for people with motor disabilities). One option is to explore other mental strategies for BCI control. A broad set of possibilities is proposed in the literature, such as imagining music 9 , phoneme imagery 10 , visual imagery of faces 11 , mental rotation and word association 4. Here we are interested in mental tasks that have not been much tested yet for controlling a BCI, namely visual imagery of objects. We focus on visualization in this work because we respond to and process visual data better than any other type of data: the human brain processes images 60,000 times faster than text 12 , and 90 percent of information transmitted to the brain is visual 13. Since we are visual by nature, and given how important the visual modality is, we propose to use this skill to design control strategies for BCIs.