BACKGROUND: In Japan, the severity staging system for idiopathic pulmonary fibrosis (IPF) has been used to determine medical care subsidies. However, this system requires invasive procedures to measure arterial oxygen tension. Recently, noninvasive and simple measurements of oxygen saturation by pulse oximetry (SpO2) have been used for severity assessments. We propose a pulse oximetry saturation (POS) staging system consisting of SpO2 parameters to predict prognosis. METHODS: We developed four prototype staging systems based on SpO2 at rest and desaturation, and adopted the system with the highest C-statistic as the POS staging system. The cutoff SpO2 values at rest were 96% and 90%, and desaturation was defined as SpO2 < 90% at the end of the 6-min-walk test. RESULTS: Two-hundred and nineteen IPF patients were studied and the C-statistic values of models 1, 2, 3, and 4 were 0.633, 0.643, 0.630, and 0.673, respectively. We judged model 4 to be a superior POS staging system; it defined SpO2 ≥ 96% at rest without desaturation as stage Ⅰ; SpO2 ≥ 96% at rest with desaturation or SpO2 90%-95% at rest without desaturation as stage Ⅱ; and SpO2 90%-95% at rest with desaturation or SpO2 < 90% at rest as stage Ⅲ. The hazard ratios of POS stage Ⅰ, Ⅱ, and Ⅲ were 1.00, 2.25, and 4.99, respectively. The C-statistic of the POS staging system produced from 1000 bootstrap samples was similar (0.673), suggesting good internal validation. CONCLUSION: A noninvasive and simple POS staging system defined by SpO2 can easily predict prognosis.