The present work explores two different effects in orthographic processing in visual word recognition. The first part is motivated by the lexical competition hypothesis which suggests that the process of recognizing a word is mediated by competitive mechanisms between visually similar possible candidates. The lexical competition effects are explored in lexical decision studies accompanied by competitive network model simulations. The studies compare findings with the conventional masked-priming paradigm with those obtained with a modified version of this procedure, designed to decrease lexical competition effects. The results are discussed in terms of their theoretical and methodological contributions. The second part of the thesis relates to letter level processing in word recognition. It explored effects of repeated letters with the regression and the factorial approaches in combination with computational modelling methodology. The regression approach is applied to megastudy data in English, Dutch, and French. The factorial approach explores the effect across several different experimental paradigms: masked-primed lexical decision and same-different tasks as well as a two-forced choice perceptual identification task. The findings are presented along with discussions of their important implications for developing theories of letter and word processing and models of visual word recognition.