In this paper we propose a computational method for determining the orthographic similarity between Romanian and related languages. We account for etymons and cognates and we investigate not only the number of related words, but also their forms, quantifying orthographic similarities. The method we propose is adaptable to any language, as far as resources are available.