MicroRNAs (miRNAs) are an evolutionarily conserved component of gene regulatory networks and essential for the development and maintenance of all multicellular organisms. miRNAs interfere with the posttranscriptional regulation of multiple genes by patterns of Watson–Crick base pairing between miRNAs and target messenger RNAs (mRNA). In the past decade, miRNA target prediction has been considered one of the most relevant topics in miRNA research. In this review, we discuss the fundamental biological concepts and commonly adopted computational and experimental methods used for predictions and validations of miRNA targets. In recent years, there has been significant progress in miRNA target predictions due to rapid advances and popularization of genome-wide immunoprecipitation assay, RNA sequencing, machine learning, and data integration. However, most of current prediction strategies and algorithms are biased by knowledge-based heuristics from early studies almost 20 years ago and reliance on canonical Argonaute (AGO)-associated mechanisms. Therefore, to improve miRNA target prediction tools, we propose expanding the study of miRNA interactomes to include (1) multiple layers of regulatory evidence, (2) data-driven methods to propose new mechanistic hypotheses, (3) genome-wide interaction assays including RNA binding proteins other than AGOs, and (4) the combined effect of mRNA structure and epigenetic modifications. Thus, the development of integrated platforms will contribute to better characterization and understanding of miRNA interactome underpinning cell functions and dysfunction leading to diseases. Finally, downstream benefits comprise new research tools to expand fundamental knowledge on miRNA biology but also to optimize rational design of artificial miRNAs as a new class of gene therapy by reducing off-target effects while increasing binding affinity to targets of interest.