Soluble HLA-G (sHLA-G) has been reported as a biomarker for embryo quality in human in vitro fertilization (IVF). The first report showed that some embryos had secreted sHLA-G and that the secretion was necessary for implantation. If these data are true, sHLA-G could be a very useful marker for embryo selection. However, one major limitation of the report was a distinct lack of controls relevant to the methods of measurement. Therefore, we re-examined the detection of sHLA-G in 109 culture supernatants from embryos fertilized in vitro. We were unable to detect sHLA-G in any supernatant despite using a more sensitive method than that used in the previous report. We further explored other reports that had shown positive data of sHLA-G secretion from embryos, and found that neither standard curves to enable calculation of sHLA-G concentration correctly nor any data for calibration were available in these reports. In this review, we point out several problems with the detection of sHLA-G in the previous reports, and describe controls and methods that can be used to determine sHLA-G concentrations accurately.