Conventional testing can underestimate a student’s learning ability. Sociocultural barriers, test anxiety and differences in rates of brain development can skew results. In principle, neuroscience tools that allow observation of brain activity might better characterize learning gains. But neuroscience and education have not always formed fruitful partnerships. New research that more profitably pairs the two fields suggests that when students use spatial skills in the classroom, the benefits extend beyond spatial understanding to other kinds of thinking, such as mentally deciphering a problem using words. The results, published on August 10 in Science Advances, also show that neuroscience tools in a real-world classroom might predict successful learning better than more traditional approaches, such as testing and grades. This information could support programs and methods that offer the biggest learning bang for the buck.