Abstract: A novel scroll-window recursive subspace identification algorithm for a closed-loop system is proposed based on orthogonal projection (QR), which introduces QR decomposition to implement the QR and eigenvalue decompositon and achieve a singular value decomposition. In addition, a bona fide recursive algorithm is used to update the QR decomposition. Then, two different projection variables W₁ and W₂ are considered according to the selection principle, and two recursive subspace identification algorithms based on QR are derived and proposed. Finally, the numerical simulation results show that the proposed method is efficient and feasible.