Background: The longitudinal association between persistent moderate to severe pain and subsequent long-term cognitive decline remains inconclusive. Methods: Study population came from the English Longitudinal Study of Ageing, an ongoing prospective and nationally representative cohort of community-dwelling adults aged ≥50 years in England. At waves 1 (2002/2003) and 2 (2004/2005) of the study, pain severity was measured based on pain intensity scores ranged from 0 to 10. We defined moderate to severe pain as pain intensity scores ≥5 points. Persistent moderate to severe pain was defined as participants reported moderate to severe pain at both waves 1 and 2. Standardized global cognitive Z scores derived from verbal memory, temporal orientation and semantic fluency were used as the primary outcome. Results: A total of 6,869 individuals (3,896 women; mean age: 63.9 ± 9.5 years) who have accepted twice measurements of pain at waves 1 and 2 (baseline), and at least one reassessment of cognitive function at waves 3 (2006/2007) to 8 (2016/2017), were included in this study. Each 5-point increase in the sum of pain intensity scores was associated with a faster rate of -0.009 (95% CI: -0.013 to -0.006, p < .001) in global cognitive Z scores. Compared with no pain group, persistent moderate to severe pain group was associated with a significantly faster decline rate of -0.031 SD/year (95% CI: -0.043 to -0.018), in global cognitive Z scores. The relationships of persistent moderate to severe pain with verbal memory, temporal orientation and semantic fluency were similar. Conclusion: Cognitive function should be monitored in individuals with persistent moderate to severe pain.