Transplantation of human embryonic stem cells (hESC) into immune-deficient mice leads to the formation of differentiated tumors comprising all three germ layers, resembling spontaneous human teratomas. Teratoma assays are considered the gold standard for demonstrating differentiation potential of pluripotent hESC and hold promise as a standard for assessing safety among hESC-derived cell populations intended for therapeutic applications. We tested the potency of teratoma formation in seven anatomical transplantation locations (kidney capsule, muscle, subcutaneous space, peritoneal cavity, testis, liver, epididymal fat pad) in SCID mice with and without addition of Matrigel, and found that intramuscular teratoma formation was the most experimentally convenient, reproducible, and quantifiable. In the same experimental setting, we compared undifferentiated hESC and differentiated populations enriched for either beating cardiomyocytes or definitive endoderm derivatives (insulin-secreting beta cells), and showed that all cell preparations rapidly formed teratomas with varying percentages of mesoderm, ectoderm, and endoderm. In limiting dilution experiments, we found that as little as two hESC colonies spiked into feeder fibroblasts produced a teratoma, while a more rigorous single-cell titration achieved a detection limit of 1/4000. In summary, we established core parameters essential for facilitating safety profiling of hESC-derived products for future therapeutic applications.