The Swedish Armed Forces and the Swedish Defence Materiel Administration are planning to start using pRFID (passive Radio-Frequency IDentification) in the material handling process. In the first step, personal equipment like clothing, knives and sleeping bags will be marked. A pRFID system consists of a reader and tags. The tags does not require an own energy source, as they are activated and powered by the energy transmitted by the reader. The types of tags considered by the Swedish Armed Forces has a reading range up to a few meters under normal circumstances. If personal equipment is to be tagged permanently with pRFID tags, the tags needs to be safe, secure and robust. The purpose with this work here is to adress a number of potential risks with using pRFID for marking personal equipment. This is studied by literature studies, calculation, and measurements and practical tests. The results show that pRFID tag will not be a problem during mine search. The tags are also deemed to be robust in pulsed electromagnetical fields, for example from radars, and the risk of tags catching fire due to exposure to this kind of fields is deemed to be very small. Nor is the transmitting power from reader and tag deemed to be above the allowed levels for human exposure. A RFID reader could in the worst case cause interference problems to electrical equipment if it is located closely to the reader, up to a few meters. In the intended application at the Swedish Armed Forces however, it should nor be a problem to avoid placing sensitive equipment close to the reader. Theoretically, the reader can also interfere with radio receivers at up to 80 m distance. This distance will be decreased decreased due to pathloss from obstacles like buildings. In practice, the risk will be low, as RFID systems have low duty cycle. Calculations show that there is a risk that a pRFID tag can be activated and read at distances of up to tens of meters. This distance is likely too short for illegitimate monitoring by UAVs. Monitoring of the number of people in a building is likely not possible due to pathloss from walls and such. Listening to the reply of an already activated tag, i.e. when a tag is read by a legitimate reader, will likely be possible at considerably larger distances. This implies that there is a technical risk that information from store-rooms, loading of transports, and so on is monitored.Earlier works have shown that it is possible to clone a tag, or replay a recorded answer from a tag in order to deceive the reader ("spoofing"). There are functions in the standard EPCGlobal Class 1 Generation 2 version 2 that are intended to protect against this, but they are optional to implement. Studying how well they work have not been part of this study.