An investigation of fatigue crack propagation behavior in air of as-deposited electroslag welds in two hot-rolled structural steels, prepared according to ASTM Specification for Structural Steel (A 36) and ASTM Specification for High-Strength Low-Alloy Structural Steel with 50 000 psi Minimum Yield Point to 4 in. Thick (Grade A) (A 588), has shown that the crack-growth rates (da/dN) in the welds were similar to or up to five times slower than the rate in the base steels. The retardation in crack-growth rate was observed to be substantially greater for crack propagation in the coarse-grained heat-affected-zone (HAZ) and bond-line regions than in the weld metal, but this was not consistently observed. Moreover, marked variations in the microstructure of the weld metal and HAZ did not significantly influence crack propagation. These crack-growth data for the weldments, described in an earlier paper, have been analyzed in the present paper in terms of residual stresses causing the observed retardation in crack-growth rate. In the present work, quantitative comparisons of the crack-growth behavior of the welds were based on a cyclic-life parameter (N₁), which was defined as the number of cycles required for precracking the IT wedge-opening-loading (WOL) specimens to a total crack length of 2.54 cm (1 in.). This parameter gave results that were consistent with those based on incremental crack-growth (da/dN) data in the earlier paper. The observed retardation in crack-growth rate in the welds is attributed to compressive residual stresses introduced by welding. This effect was analyzed in the present paper in terms of a stress-intensity-range suppression concept, whereby the applied stress-intensity-factor range is decreased to some lower effective value. The results showed that the beneficial effect of compressive residual stresses on crack propagation appears to be of a variable nature and would diminish following a stress-relief heat treatment.