This study investigates the possibility and efficacy of paper-based, in-class, data-driven learning (DDL) of academic lexical bundles below the C1 level of proficiency described by the Common European Framework of Reference (CEFR; advanced high ACTFL). A two-stage experimental design involving three groups (n = 41) and 24 two-to-four word academic items was implemented. First, the question of whether this type of learning works with these items below the C1 level is addressed through a nonequivalent-groups quasi-experimental design covering a five-week period. The results indicate that this technique is effective at the B2 level, but not at the A2-B1 level. Next, an equivalent-groups experimental design compares this style of learning to conventional techniques at the B2 level. The results of this stage suggest that paper-based, in-class DDL is more effective than conventional learning with academic lexical bundles at the B2 level.