Pseudowords (PW) are strings of phonemes/graphemes that conform to the phonological and orthographic rules of the language but carry no meaning. These properties make them suitable for testing phonological processing in speech-language therapy assessment, but also for the improvement of this ability during intervention. The motivation for this study stems from the lack of guidelines for forming a list of PW, as well as from the lack of systematic research on inherent properties of PW in Croatian. For this reason, the first part of the paper presents the method of PW formation, taking into account the peculiarities of the syllabic and phonetic system Croatian, following the approach of combining subsyllabic elements. In the second part of the paper the correlation between success in repetition and reading of newly formed PW and their phonological and lexical features was investigated in beginning readers and children who are in the stage of reading automatization. Partly consistent with the prediction, a significant moderate to high correlation was found between success in repetition and reading PW, and linguistic features (length, segmental complexity, and wordlikeness) in both groups. The only exception is a non-significant correlation between segmental complexity and success in PW repetition in children who are in the automatization stage of reading. It appears that as children grow older and are exposed to different phonological combinations, they integrate different linguistic information and prosodic features without difficulty, which is still challenging for beginning readers. Clinical implication of the study is that it provides knowledge of PW formation, to ensure their validity in intervention procedures.

pseudowords, pseudoword formation, inherent linguistic features, speech-language intervention, phonological processing