BACKGROUND: Longitudinal assessments of psychological effects related to length and recency of quarantine experience in general populations are of importance. We aim to investigate if recency and duration of quarantine exposures during the COVID-19 pandemic were associated with mental health, across subgroups. METHODS: We included three prospective cohorts from Iceland and Norway with data on quarantine and symptoms of depression and anxiety from March 2020 to March 2022. We calculated prevalence ratios (PR) of probable depression and anxiety in relation to quarantine exposure, and performed longitudinal analyses in a subpopulation with repeated assessments to test the potential change in mental health burden due to quarantine over time while controlling for current quarantine status and other covariates. RESULTS: In total, 105,344 and 94,435 individuals were included in the analysis of probable depression and anxiety, respectively, with 18.2% and 40.0% reporting quarantine exposure before the most recent assessment of corresponding mental health symptoms. Overall, quarantine exposure was associated with probable depression (PR 1.19 [95% CI: 0.99-1.42]) and anxiety (PR 1.21 [1.08-1.36]). Compared to individuals without quarantine, being exposed to quarantine for 0-2, 2-4, or > 4 weeks was associated with incrementally higher prevalence of probable depression (PR 1.15 [0.93-1.43]; 1.34 [1.06-1.68]; 1.72 [1.35-2.18], respectively) and probable anxiety (PR 1.12 [1.01-1.23]; 1.28 [1.14-1.45]; 1.76 [1.56-1.98]) in a step-wise manner; those who were quarantined within the last 2 weeks, last 2-4 weeks, or earlier showed a higher prevalence of probable depression in a dose-response manner (PR 1.62 [1.32-1.99], 1.32 [1.07-1.63], and 1.23 [1.06-1.43], respectively). The prevalence of probable anxiety did not appear to differ by the recency of quarantine. The longitudinal analyses (mean follow-up: 20.5 months) confirmed significantly higher prevalence of probable depression but only among those who were quarantined for > 4 weeks (PR 1.61 [1.30-2.00]), and of probable anxiety among those quarantined 2-4 weeks (PR 1.29 [1.14-1.45]) and > 4 weeks (PR 1.56 [1.34-1.82]). CONCLUSIONS: This study underscores the importance of monitoring mental well-being of populations recently quarantined, particularly those quarantined for prolonged periods. Greater emphasis should be placed on the detrimental psychological effects in the risk-cost-benefit analysis of quarantine as a mitigation strategy in future pandemics.