Adhesive joints in real-world conditions often experience variable or step loading rather than constant-amplitude fatigue. This study addresses this gap by examining the influence of load sequence and block loading on fatigue damage in adhesive joints of fiber-reinforced polymer (FRP) composites. A novel bilinear traction-separation law based on the Fatigue Crack Growth Rate (FCGR) rule is introduced to predict fatigue failure under step/variable loads, accounting for load history, sequence, and interaction effects. This model was validated using a double-lap joint model under step/variable loading across four experimental scenarios. The proposed model outperformed existing fatigue damage-accumulation models, significantly reducing the Relative Error of Prediction (REP). Notably, the proposed model significantly reduced the Relative Error of Prediction (REP), achieving reductions from 81.10% to as low as 0.013% in certain cases. The proposed bilinear law exhibited an accelerated damage accumulation rate per cycle for low-to-high loading situations and a decelerated rate for high-to-low loading scenarios, aligning more closely with experimental observations. The proposed model offers practical benefits by improving fatigue life predictions, enabling optimized FRP composite designs, and minimizing overengineering. These advancements are particularly relevant in industries such as aerospace, automotive, and wind energy, where structural durability and safety are paramount. This research represents a significant step forward in the fatigue analysis of composite adhesive joints, paving the way for more reliable engineering solutions.