In the present study, we investigated the facilitation effects of lexical-semantic preactivation during interclause processing, in a sample of 22 Spanish L1 undergraduate students. Previous research has mainly examined nouns’ predictability effects at clause level and focused on semantic factors that precede text interpretation. Yet, it remains unclear if such effects occur when text interpretation is not only preceded but also fed into, for instance, by subjective clause adverbs expressing positive/negative polarity in Spanish. Drawing on a self-paced reading task, we measured response times at the target nouns and fitted mixed-effects models on the outcome variable to the targets. We observed longer response times at low predictable nouns when negative adverb polarity mediated interclause processing. This finding suggests that predictability effects may be detected at early processing cycles, even before the noun introducing the lexical-semantic update is perceived as current input, thus facilitating the construction of a coherent text model.