In a word-by-word, self-paced reading experiment, we examined the effect of voice and plausibility on reading time, as well as accuracy and reaction time to a comprehension question. The effect of the interaction of verbal IQ with these factors was also explored, and comparisons were made between trials interpreted correctly and incorrectly. A linear mixed effects regression (LMER) analysis showed that performance was impaired in all three measures in response to an implausible sentence, but voice only had an effect in offline measures. A higher verbal IQ lessened participants’ susceptibility to the effects of implausibility. In trials where participants misinterpreted the sentence, implausibility had no effect on reading time, whereas participants slowed down in response to implausibility in correctly interpreted sentences. In contrast, slower reaction times were observed in misinterpreted trials than in correctly answered ones. We discuss these results in terms of algorithmic and heuristic processing, and compare our findings with Townsend and Bever’s (2001) LAST. Suggestions for further research are also made.