The present study examines the relationship between learner participation and accuracy improvement in the use of some features of standard Italian, which were taught over a period of intensive grammatical instruction in a tertiary classroom setting. The subjects of the study are 60 tertiary learners of L2 Italian from a variety of Italian dialect backgrounds, aged 18-25, receiving instruction focused on the use of the past tense system. The incidence of learner overtly vocal participation is measured by monitoring the number of verbalised exchanges of each participant over the 15 hours of instruction. Less overtly vocal participation is monitored by teacher observation of student interaction. Data relating to both vocal and less vocal participation are compared to the improvement percentages in the accurate use of the taught features. While the results of the investigation do not support unambiguously the hypothesis that measurable, verbalised participation during instruction relates to higher accuracy achievement, they bring to the foreground the crucial role of less overtly vocal manifestations of learner participation. This calls for a re-definition of the notion of learner participation and has pedagogical implications for a common assumption held by the L2 teaching profession. (Contains 2 tables.)