We examine orthographic learning and its variation across four languages with differing orthographic depths: English, German, Spanish, and Turkish. Orthographic word learning refers to individuals acquiring novel words’ spellings and storing their representations in their mental dictionary (Ehri, 2013). We designed an experimental learning task, where participants learn the spellings of novel pseudowords. To understand what influences orthographic learning, we manipulated orthographic depth, which reflects how consistently a word’s letters map onto sounds in a given language (Katz & Frost, 1992), and orthographic typicality, how closely a word’s spelling matches the typical letter combinations in the target language (Woollams et al., 2010). The participants were 186 adults who were native speakers of English, German, Spanish or Turkish, which vary in orthographic depth. We presented 40 pseudowords to our participants, with each seeing either 20 novel items or their 20 homophone pairs (words that are pronounced the same but spelled differently). Half of the items were spelled according to the orthographies’ spelling conventions, and the other half had atypical spellings. We found no effects of orthographic depth or typicality; however, similarity to real words in the respective languages strongly predicted performance across tasks. This study was the first to compare languages with different orthographic depths using identical pseudowords while including typicality and orthographic similarity, providing a framework for future cross-linguistic work.