To determine the fracture resistance and its constraint effect for storage tank A285 carbon steel, a set of single edge notched bend (SENB) specimens with different crack lengths are tested under the guideline of ASTM standard E1820. Normalization method is adopted to develop J–R curves by directly using the load versus load-line displacement data without the need for online crack size measurement. For validation, the conventional electric potential drop method is also used to monitor crack extension, and to determine the J–R curves. The results indicate that the crack extension, J–R curves and fracture initiation toughness determined from both methods are nearly identical for the SENB specimens considered. The J–R curves obtained from the A285 SENB specimens show a strong dependence on crack size or fracture constraint. In reference to the J–A 2 fracture theory, the constraint effect is quantified by a constraint parameter A 2. Through experimental and numerical analyses of the specimens, a constraint-corrected J–R curve is obtained as a function of crack extension and the constraint parameter A 2. It is concluded that the normalization method is valid and cost-effective for the determination of J–R curves from both standard and nonstandard SENB specimens. In addition, the parameter A 2 can reasonably quantify the constraint effect on the experimental J–R curves.