Visual information can be processed by humans both consciously and unconsciously. Under conscious conditions the humans are aware of the input, and thus can process it according to their goals. However in unconscious conditions participants remain unaware of the input, as evident from their chance accuracy when judging about the input. It has been shown that although participants may not consciously perceive subliminal input, the input can be processed automatically. Masked priming is an experimental paradigm to investigate unconscious visual processing. In this paradigm the input is presented for a very brief duration, often temporally sandwiched between a pre- and a post-mask. In this thesis, I and we mostly applied the masked-priming paradigm to study human unconscious visual processing using words and face images as input. We conducted five studies to understand unconscious visual processing and compare it with conscious processing. Our first study is about the accurate measurement of the visibility of subliminally presented visual input. In the second study, we provide a test-case for the embodied cognition theory using subliminal words. The third study is concerned with conflict control under aware and unaware conditions. In our fourth study, we tried to understand the role of subcortical structures in subliminal face processing. The fifth study is concerned with relations between space and valence priming effects. Together our findings further the understanding of unconscious and conscious visual processing.