Fatigue usage monitoring systems (Fums) offer considerable potential for life extension of aircraft parts. In this work the life extension benefits of Fums is assessed by adopting a probabilistic approach. The roles of damage law type and of service usage variability is explored. It is shown by analysis that in the absence of cycle to cycle load interaction effects, load sequence has no effect on eventual life in either linear or non linear damage laws, provided that the function describing the rate of damage growth has separable variables of stress and damage. This condition includes fracture mechanics crack growth laws. Monte Carlo simulations have been conducted of fatigue life distributions in helicopter rotor components. Variability in manoeuvre damage, when summed over a large number of manoeuvres, has little effect on scatter in overall lives. A fixed manoeuvre usage spectrum will result in very small scatter in lives, whereas keeping the usage constant for each helicopter and allowing it to vary between helicopters, produces a significantly increased variability. The influence of load factors on life is also assessed. The extent of possible maintenance credits is discussed together with the use of Bayesian updating to make optimum use of both prior design data and current loads or damage information provided by Fums.