Objective To explore the effects of physical activity (PA) on lipid profile and identify the optimal intensity of PA that improves dyslipidemia. Methods Community-based individuals aged 40-60 years old were recruited by a cluster sampling method. PA was estimated using the International Physical Activity Questionnaire, and PA levels of individuals were classified as low, moderate, or high. Lipid profiles were measured by using fasting blood samples, and the potential confounding variables related to lipid profiles were comprehensively collected. Unconditional logistic regression method was used to investigate the relationship of PA with lipid profiles. Results A total of 5 664 subjects (38.37% men) were included in this study. The percentages of individuals with low, moderate and high PA levels were 9.99% (566/5 664), 44.79%(2 537/5 664) and 45.22%(2 561/5 664), respectively. High PA could reduce the risk of dyslipidemia, with an odds ratio (OR) of 0.83 (95% confidence interval [CI]: 0.71-0.98) for dyslipidemia compared to low PA, while the association was not found for moderate PA. High level PA mainly influenced the high-density lipoprotein cholesterol and triglycerides. Conclusion High level PA, compared with low level, can reduce the risk of dyslipidemia in middle-aged Chinese population.