Due to changes in global security requirements attention is turning to new means by which anomalies on the human body might be identified. For security screening systems operating in the millimeter wave band anomalies can be identified by measuring the emissivities of subjects. As the interaction of millimeter waves with the human body is only a fraction of a millimeter into the skin and clothing has a small, but known effect, precise measurement of the emission and reflection of this radiation will allow comparisons with the norm for that region of the body and person category. A technique to measure the human skin emissivity in vivo over the frequency band 80 GHz to 100 GHz is developed and described. The mean emissivity values of the skin of a sample of 60 healthy participants (36 males and 24 females) measured using a 90GHz calibrated radiometer were found to range from 0.17±0.005 to 0.68±0.005. The lower values of emissivity are a result of measuring particularly thin skin on the inner wrist, volar side of the forearm, and back of hand, whereas higher values of emissivity are results of measuring thick skin on the outer wrist, dorsal surface of the forearm, and palm of hand. The mean differences in the emissivity between Asian and European male participants were calculated to be in the range of 0.04 to 0.11 over all measurement locations. Experimental measurements of the emissivity for male and female participants having normal and high body mass index indicate that the mean differences in the emissivity are in the range of 0.05 to 0.15 for all measurement locations. These results show the quantitative variations in the skin emissivity between locations, gender, and individuals. The mean differences in the emissivity values between dry and wet skin on the palm of hand and back of hand regions were found to be 0.143 and 0.066 respectively. These results confirm that radiometry can, as a non-contact method, identify surfaces attached to the human skin in tens of seconds. These results indicate a route to machine anomaly detection that may increase the through-put speed, the detection probabilities and reduce the false alarm rates in security screening portals.