The family plays a vital role in fostering children's learning in science through joint experiences in diverse settings such as homes or museums. Beyond frequency, the quality of parent–child interactions in science significantly influences the children's development. However, research in this area has often focused on single aspects of parent–child interactions, exhibiting substantial variation in operationalisations and terminologies across studies. This complicates the comparability between the studies and an identification of which elements in high-quality parent–child interactions are especially important for children's science learning. To address this gap, we systematically reviewed 81 studies on shared parent–child interactions in science, and organised their operationalisations within four overarching categories of interaction quality: parents' science-specific support, cognitive activation, cognitive support, and emotional-motivational support. Most studies placed emphasis on parents' provision of science-specific support and cognitive activation, with less attention given to cognitive and emotional-motivational support. Further, we found mixed findings on how interaction quality relates to three aspects in children's science learning: their cognitive and motivational-affective outcomes, and engagement. These inconsistencies may stem from variations in operationalisations and measurements across studies, and individual prerequisites in children. Overall, this review may improve the comparability between the studies and identification of blind spots in research. It may aid future research in clearer operationalisation and recognition of boundaries in assessing interaction quality, potentially leading to more generalisable findings. Additionally, this review highlights the need for future research to include families from more diverse cultural or social backgrounds and adopt a broader perspective on parent–child interaction quality.