Visual masking and visual suppression by transcranial magnetic stimulation (TMS) are both widely utilized in cognitive neuroscience to investigate a wide range of processes. However, the neural processes affected by visual masking and TMS remain unclear. We compared para- and metacontrast masking with TMS-induced suppression of visibility in a within-subjects design where participants were asked to detect and rate the visibility of a stimulus. TMS pulses applied 75-109 msec after the onset of the visual stimulus reduced the subjective visibility of the target. Even when the TMS pulses completely eliminated the conscious perception of the target, unconscious location detection was possible. The visual masking condition yielded similar results: metacontrast did not eliminate unconscious location detection even when the target was reported not seen at all. As the first target-related signals were likely to reach the visual cortex before TMS pulses started to modulate target visibility, we suggest that TMS and metacontrast masking affected neural signals subsequent to the target's transient onset-response. This implies that a preserved onset-response is sufficient for unconscious processing of stimulus attributes, but not for conscious perception. (C) 2010 Elsevier Srl. All rights reserved.