Both probability of default (PD) and loss given default (LGD) constitute relevant input parameters for credit risk management in pillars I and II. Assuming that both default data and loss data have been successfully collected and that PD and LGD estimators have been successfully obtained afterward, both parameters have to be validated in the following period(s). Whereas PD validation techniques have been intensively discussed in the past decade, quantitative LGD validation has not received sufficient attention in the literature, so far. In this light, the focus of this work is to summarize and classify, analyze and compare possible LGD validation instruments. Beyond that, possible extensions are introduced and discussed, and application is given to a hypothetical LGD data set.