In the course of inquiry activities similar to those of real scientists, learners are supposed to develop knowledge both on the level of observable phenomena and on the level of explanatory theories. However, some theories involve theoretical entities (e.g., "Weiss domains") that cannot be observed directly and therefore may be hard to discover. Presenting theoretical ideas before or after inquiry activities might help to overcome these difficulties. However, whereas prior presentation of theoretical ideas enables learners to apply these ideas during inquiry activities and may therefore have a sustainable effect on knowledge on the theory level, subsequent presentation of theoretical ideas does not allow for applying these ideas during inquiry activities and may therefore fail to yield a lasting effect on knowledge on the theory level. In contrast, specific scaffolds designed to guide learners during their inquiry activities to foster knowledge on the level of phenomena may be ineffective with respect to knowledge on the theory level. A 2?×?2?×?2-factorial experiment with the factors "specificity of scaffolds" (unspecific/specific), "prior presentation of theoretical ideas" (no/yes), and "subsequent presentation of theoretical ideas" (no/yes) was conducted in an inquiry unit about magnetism. The sample analyzed comprised 538 pupils from 23 seventh-grade classes. In an immediate and a delayed posttest 2 months later, both knowledge on the level of phenomena and knowledge on the theory level were measured. Prior presentation of theoretical ideas had an immediate as well as a longer-term effect on knowledge on the theory level, whereas subsequent presentation of theoretical ideas only had a short-term effect. Specific scaffolds had no effect on any of the outcome variables. These findings suggest that learners may in fact derive predictions and explanations of the outcomes of their experiments if theoretical ideas are presented beforehand, and thereby develop deeper theoretical understanding.