Abstract GWAS have identified thousands of SNPs associated with different human phenotypes. Most SNPs are located in non-coding regions of the genome, where a small subset of causal SNPs is thought to contribute to risk by regulating transcription. However, identification of such causal variants remains challenging due to linkage disequilibrium, where blocks of SNPs are inherited together. Here, we perform a massively parallel reporter assay called SuRE in human neural stem cells to assess the transcriptional regulatory potential of 7.1 million SNPs. We identify 7,002 SNPs with regulatory potential (emVars) among the 484,057 SNPs with sufficiently high expression. These emVars are enriched in enhancers and often coincide with binding sites for transcription factors, such as ZBTB33, YY1, and ETS factors. We overlap SuRE data with eQTL data to pinpoint likely gene-regulatory variants. Additionally, combining SuRE data with GWAS for seven neuropsychiatric phenotypes identifies 187 emVars in risk loci containing 183,186 SNPs assessed by SuRE. In conclusion, our SuRE dataset provides a rich source to prioritize genetic variants that may be relevant for complex human phenotypes.