Abstract: Cancer cells have been found to frequently express genes that are normally restricted to the testis, often referred to as cancer/testis (CT) antigens or genes^{1, 2}. Because germ cell specific antigens are not recognized as “self” by the innate immune system³, CT-genes have previously been suggested as ideal candidate targets for cancer therapy⁴. The use of CT- genes in cancer therapy has thus far been unsuccessful, most likely because their identification has relied on gene expression in whole testis, including the testicular somatic cells, precluding the detection of true germ cell specific genes. By comparing the transcriptomes of micro-dissected germ cell subtypes, representing the main developmental stages of human spermatogenesis⁵, with the publicly accessible transcriptomes of 2.617 samples from 49 different healthy somatic tissues⁶ and 9.232 samples from 33 tumor types⁷, we here discover hundreds of true germ cell specific cancer expressed genes. Strikingly, we found these germ cell cancer genes (GC-genes) to be widely expressed in all analyzed tumors. Many GC-genes appeared to be involved in processes that are likely to actively promote tumor viability, proliferation and metastasis. Targeting these true GC-genes thus has the potential to inhibit tumor growth with infertility being the only possible side effect. Moreover, we identified a subset of GC-genes that are not expressed in spermatogonial stem cells. Targeting of this GC-gene subset is predicted to only lead to temporary infertility, as untargeted spermatogonial stem cells can recover spermatogenesis after treatment. Our GC- gene dataset enables improved understanding of tumor biology and provides multiple novel targets for cancer treatment.