A thorough understanding of lexical access is fundamental for speech perception models, and the word reconstruction task helps in achieving that. In a word reconstruction task, participants are presented with nonwords and instructed to change those into real words by changing one phoneme. Previous research consistently shows a C-bias across languages, meaning participants are more likely to change a vowel phoneme than a consonant phoneme, indicating a reliance on consonants for word recognition and pointing to an intrinsic difference between these phoneme types. However, these studies only used isolated words to investigate this difference. Given that speech occurs as both isolated words and in sentences, and sentential context influences perception, this study investigated how placing nonwords in sentences affects the roles of consonants and vowels in Dutch word recognition. Participants (N = 37) were divided into either the sentences or isolation condition. The hypothesis was that sentential context would alter the C-bias, though the direction was undefined. GLMM analysis revealed no significant difference in how nonwords were perceived between isolation and sentences conditions. Unexpectedly, no C-bias was found in the isolation condition, unlike prior research. Participants indicated that the stimuli could not always be clearly heard, affecting their answers. Optimizing clarity of stimuli is thus essential for future studies. This study was a starting point for research on the effect of sentential context on speech perception, at the level of consonants and vowels. Any such effect should be integrated into speech perception models. Thus far, no significant effects were found.