This article highlights the need for a new type of corpus that will combine both a learner and a pedagogic corpus. The need for such a type of corpus is underlined by the interaction between input and output and the expected impact of the former on the latter within all instructional settings. Such a corpus will allow us to collect data, including teacher language, learner language, coursebook content, etc. and relate it to students' output at specific points in time. Our study aims to present and describe the "Young Learner Corpus of English" (henceforth, "YoLeCorE"), which is an example of a combined pedagogic and learner corpus. "YoLeCorE" consists of 1,5 million tokens and includes both the input and the output of 17 Greek young learners (aged 8-9) of English as a foreign language. "YoLeCorE"'s unique characteristic is that it includes not mere samples of learner language, but all the language that each and every one of the students listened to, read, or produced -- orally and in writing -- during a whole school year. It includes four distinct sub-corpora (one for each language skill) and allows researchers to conduct a variety of analyses and thus track the process of instructed second language acquisition.