The objective of this study was to measure the threshold stress-intensity factor at crack arrest (KTHa) for commercial Cr-Mo and Ni-Cr-Mo pressure vessel steels exposed to 100 MPa hydrogen gas. The primary aim was to evaluate the effects of material variables, such as yield strength, on the measured KTHa. Another goal was to assess testing procedures to identify any variables that may bias the measurements. The measured KTHa decreased as a function of increasing yield strength, consistent with the well-known effect of strength on hydrogen-assisted fracture. Unexpectedly, the measured KTHa increased as the initial applied stress-intensity factor (Kapp) increased. This trend did not result from any compromise in the crack-tip mechanics associated with the dimensions and mechanical loading conditions of the test specimens. Rather, it is hypothesized that the KTHa vs Kapp relationship resulted from the transition of the crack-tip strain field from the stationary-crack limit toward the propagating-crack limit as subcritical cracking proceeded. This transition in the crack-tip strain field was significant for the following reasons: 1) hydrogen-assisted, subcritical cracking in the pressure vessel steels was strain-driven, and 2) the strain field characteristic of stationary cracks was more intense compared to the strain field associated with propagating cracks. The KTHa vs Kapp trend reflected the interplay between the strain field vs crack length relationship and the K vs crack length relationship (which depended on Kapp).