Genotype imputation is a technique extensively used in genome-wide association studies (GWAS), as well as in their meta-analysis. In this talk, we outline strategies to account for imputation accuracy when including imputation-based GWAS results in a meta-analysis, considering both fixed-effect and random-effects models. While adding studies to the meta-analysis typically boosts the power to detect genetic associations, inclusion of imputation-based studies may adversely affect power due to increased genotype uncertainty. We address this trade-off via a reweighing scheme that controls the contribution of imputation-based studies in meta-analyses. The proposed method achieves a better detection power relative to traditional approaches, and improves the validity and reliability of imputation-based meta-analysis results.