The paper addresses the identification of Hammerstein systems under unknown periodic disturbances, focusing on improving the accuracy and reducing the variance of parameter estimates. The existing over-parameterized model-based subspace identification (OPM-like SIM) methods, while effective, require estimating extra parameters and performing a low-rank approximation, which can lead to high variance, especially with small and noisy data sets. To overcome this, the author proposes a parsimonious model-based SIM that uses two parsimonious models to describe the Hammerstein system and introduces an orthogonal projection-based fixed-point iteration method to eliminate disturbance effects and provide consistent parameter estimates. This approach avoids the need for extra parameter estimation and low-rank approximation, potentially enhancing the accuracy and variance properties of the estimates. The effectiveness of the proposed method is demonstrated through mathematical proofs and simulation examples. The paper is structured to include problem formulation, method presentation, performance evaluation, and conclusions.The paper addresses the identification of Hammerstein systems under unknown periodic disturbances, focusing on improving the accuracy and reducing the variance of parameter estimates. The existing over-parameterized model-based subspace identification (OPM-like SIM) methods, while effective, require estimating extra parameters and performing a low-rank approximation, which can lead to high variance, especially with small and noisy data sets. To overcome this, the author proposes a parsimonious model-based SIM that uses two parsimonious models to describe the Hammerstein system and introduces an orthogonal projection-based fixed-point iteration method to eliminate disturbance effects and provide consistent parameter estimates. This approach avoids the need for extra parameter estimation and low-rank approximation, potentially enhancing the accuracy and variance properties of the estimates. The effectiveness of the proposed method is demonstrated through mathematical proofs and simulation examples. The paper is structured to include problem formulation, method presentation, performance evaluation, and conclusions.