The mean time to failure of a device, component, or any part or subsystem that can be separately tested. For any particular item the mean life will be based on the results of life tests and one or more of the following will be quoted.

The observed mean life is the mean value of the observed times to failure of all specimens in a sample of items under stated stress conditions. The criteria for what constitutes failure should be stated.

The assessed mean life is the mean life of an item determined as a limiting value of the confidence interval with a stated probability level based on the same data as the observed mean life of nominally identical items. The following conditions apply: the source of the data should be stated; results may be combined only when all conditions are similar; it should be stated whether one- or two-sided intervals are being used; the lower limiting value is usually used for mean life statements and the assumed underlying distribution should be stated.

The extrapolated mean life is the extension of the observed or assessed mean life by a defined extrapolation or interpolation for stress conditions different from those applying to the conditions of the assessed mean life.

The elapsed time at which a stated proportion (q per cent) of a sample or population of items has failed is the q-percentile life. This is quoted as the observed percentile life, assessed percentile life, and extrapolated percentile life under similar conditions to those quoted for mean life.