When missing values occur in multi-view data, all features in a view are likely to be missing simultaneously. This leads to very large quantities of missing data which, especially when combined with high-dimensionality, makes the application of conditional imputation methods computationally infeasible. We introduce a new meta-learning imputation method based on stacked penalized logistic regression (StaPLR), which performs imputation in a dimension-reduced space. We evaluate the new imputation method with several imputation algorithms using simulations. The results show that meta-level imputation of missing values leads to good results at a much lower computational cost, and makes the use of advanced imputation algorithms such as missForest and predictive mean matching possible in settings where they would otherwise be computationally infeasible.