Objective: Regular physical activity (PA) is associated with less self-reported anxiety, but prospective studies linked to clinician diagnoses of anxiety disorder remain scarce. We examined whether the PA levels recommended for general health are related to anxiety symptoms and disorders.Method: In total, 43,863 Swedish adults were surveyed in 1997 and responses linked to medical registers until 2010. Weekly durations of habitual moderate and vigorous leisure time PA were self-reported. Cross-sectional and prospective relationships between the total duration (minutes) of PA, 0-149 ('below'), 150-299 ('achieve'), and >= 300 min ('exceed') with self-reported anxiety symptoms and incident anxiety disorder were explored. Associations were explored using logistic and Cox proportional hazard regression models.Results: Of 27,053 participants with complete data (mean age = 49.0 years, SD = 15.9, 66% female), 76% met the recommended duration of PA ( >= 150 min), and 38% exceeded this duration. At baseline, 2573 participants (9.5%) reported elevated anxiety symptoms. In cross-sectional analyses, engaging in >= 150 min of MVPA/week was associated with 24% (OR = 0.76, 95% CI = 0.68-0.86) lower odds of anxiety symptoms. Exceeding the weekly duration was associated with 36% (OR = 0.64, 95% CI = 0.57-0.72) lower odds. During the 13-year follow-up, 198 incident cases of anxiety disorder (0.8%) were identified. No significant prospective relationships were found.Conclusions: Engaging in leisure time PA at levels recommended for general health may reduce the risk of elevated anxiety symptoms. As the incidence of anxiety disorder was likely under-estimated, further prospective studies are needed to determine the relationship between PA and incident anxiety disorder.