miRNAs form a class of noncoding RNAs, involved in post-transcriptional regulation of gene expression, broadly studied for their involvement in physiological and pathological context. Inhibition of mature miRNA transcripts, commonly used in miRNA loss-of-function experiments, may not be specific in case of miRNAs with high sequence homology, e.g. miRNAs from the same seed family. Phenotypic effects of miRNA repression might be biased by the repression of highly similar miRNAs. Another challenge is simultaneous inhibition of multiple miRNAs encoded within policistronic clusters, potentially co-regulating common biological processes. To elucidate roles of miRNA clusters and miRNAs with high sequence homology, it is of key importance to selectively repress only the miRNAs of interest. Targeting miRNAs on genomic level with CRISPR/dCas9-based methods is an attractive alternative to blocking mature miRNAs. Yet, so far no clear guidelines on the design of CRISPR inhibition (CRISPRi) experiments, specifically for miRNA repression, have been proposed. To address this need, here we propose a strategy for effective inhibition of miRNAs and miRNA clusters using CRISPRi. We provide clues on how to approach the challenges in using CRISPR/dCas in miRNA studies, which include prediction of miRNA transcription start sites (TSSs) and the design of single guide RNAs (sgRNAs). The strategy implements three TSS prediction online tools, dedicated specifically for miRNAs: miRStart, FANTOM 5 miRNA atlas, DIANA-miRGen, and CRISPOR tool for sgRNAs design; it includes testing and selection of optimal sgRNAs. We demonstrate that compared to siRNA/shRNA-based miRNA silencing, CRISPRi improves the repression specificity for miRNAs with highly similar sequence and contribute to higher uniformity of the effects of silencing the whole miRNA clusters. This strategy may be adapted for CRISPR-mediated activation (CRISPRa) of miRNA expression.