Most microbial ecosystems cannot be understood without quantifying ecological interactions between their member species. Given the challenges of comprehensively resolving interactions experimentally, a range of prediction methods was developed. Here, we review genome-based prediction methods in particular and discuss their strengths and weaknesses. We then cover different experimental designs to explore microbial interactions and introduce methods to infer interaction signs and strengths from experimental data. Despite the range of available methods to study microbial interactions in silico and in vitro, interactions in a spatial context are still underexplored, and we lack comprehensive interaction databases, which are important gaps to fill in the future.