The case of limited data implies that some unknown uncertainties may be inv
olved in fatigue reliability analysis. For the sake of statistical convenie
nce, for consistency with the relevant physical arguments and, most importa
ntly, to ensure the safety in design evaluation, an approach is developed t
o determine an appropriate distribution, from four possible assumed distrib
utions - three-parameter Weibull, two-parameter Weibull, lognormal and extr
eme maximum-value distributions. The approach makes allowance for consisten
cy with the fatigue physics and checking tail fit effects. An application t
o nine groups of fatigue life data of 16Mn steel (Chinese steel) welded pla
te specimens shows that the lognormal distribution and the extreme maximum-
value distribution may be the appropriate distributions of the fatigue Life
under limited data. (C) 1999 Elsevier Science Ltd. All rights reserved.