This paper discusses learners’ responses to metalinguistic feedback and their strategies in error correction in a Web-based Intelligent Language Tutoring System (ILTS) for German. The system consists of a grammar and a parser which analyses student input and provides error-specific feedback to grammar and vocabulary exercises. Furthermore, the ILTS employs a Student Model that matches feedback messages to learner expertise and provides remedial exercises. Two beginner German classes, 33 students in all, used the ILTS for grammar practice, submitting a total of 4405 sentences for analysis. The results indicate that for the vast majority of sentences (79.5%) students read and attended to system feedback. Moreover, as iterations increased students paid more attention to the feedback messages.