[en] A relatively large collection of fatigue-crack growth results for annealed Types 304 and 316 stainless steels over a wide range of temperature was processed and analyzed in a consistent way. Only data that satisfied the criteria of ASTM E647-82 was retained and used in the statistical treatments that followed. Linear least-squares regression equations and 95% confidence intervals were fitted through the results for each material/temperature set. The regression results (and their associated limits of validity) provide useful equations for the analysis of structural components. Overlap (or the lack of overlap) of the confidence intervals was employed as a criterion as to whether the results for Types 304 and 316 should be separated into discrete sets, and on this basis it was concluded that the two alloys should be treated separately. 38 references, 16 figures, 1 table