Building on rich training literature, we examined which skills constituting the control-of-variables strategy (CVS) benefit from a comprehensive training, and which develop similarly during content-focused inquiry at ages 10-12. In addition, we examined whether prior knowledge, reasoning abilities, and reading comprehension explain variation in intervention effects. In a within-classroom, controlled field-experiment, half of N = 618 children from schools located in the German-speaking part of Switzerland were randomly assigned to a training on the CVS, and the other half to an active control group engaging in content-focused inquiry. Mixed-effects models revealed that the CVS training improved children's skills in planning controlled experiments and understanding the indeterminacy of confounded experiments, whereas it did not show specific effects on children's skills in identifying and interpreting controlled experiments. Children with better reasoning abilities and reading comprehension showed the strongest intervention effects on the more difficult skills. The general and differential effects of training remained mostly stable after a period of 6 months. More basic CVS skills seem to develop without targeted training, whereas more advanced ones benefit most from training that meets learners' preconditions.