Unconscious semantic priming from pictures under backward masking and continuous flash suppression

•It is debated which perceptual functions can take place unconsciously.•We measured semantic priming from invisible pictures.•No significant priming effects, despite sufficient statistical power.•Neither with backward masking nor with continuous flash suppression.•Unconscious processing may be rather limited in scope. It is debated whether the meaning of invisible pictures can be processed unconsciously. We tested whether pictures of animals or objects presented under backward masking or continuous flash suppression could prime the subsequent categorization of target words into animal or non-animal. In Experiment 1, the backward masking part failed to replicate the priming effect reported in two previous studies, despite sufficient statistical power (N = 59). Similarly, the continuous flash suppression part provided no evidence for a priming effect. In Experiment 2 (N = 65) we shortened the prime-target SOA from 290 ms to 90 ms, but again failed to obtain unconscious semantic priming under backward masking. Thus, our study did not provide evidence for unconscious semantic processing of pictures. These findings support the emerging view that unconscious processing is rather limited in scope.