Objective: To assess risk factors for persistence vs improvement and to describe clinical characteristics and diagnostic evaluation of subjects with post-acute sequelae of COVID-19/post-COVID-19 syndrome (PCS) persisting for more than one year. Design: Nested population-based case-control study. Setting: Comprehensive outpatient assessment, including neurocognitive, cardiopulmonary exercise, and laboratory testing in four university health centres in southwestern Germany (2022). Participants: PCS cases aged 18 to 65 years with (n=982) and age and sex-matched controls without PCS (n=576) according to an earlier population-based questionnaire study (six to 12 months after acute infection, phase 1) consenting to provide follow-up information and to undergo clinical diagnostic assessment (phase 2, another 8.5 months [median] after phase 1). Main outcome measures: Relative frequencies of symptoms and health problems and distribution of symptom scores and diagnostic test results between persistent cases and controls. Additional analysis included predictors of changing case or control status over time with adjustments for potentially confounding variables. Results: At the time of clinical examination (phase 2), 67.6% of the initial cases (phase 1) remained cases, whereas 78.5% of the controls continued to report no health problems related to PCS. In adjusted analyses, predictors of improvement among cases were mild acute index infection, previous full-time employment, educational status, and no specialist consultation and not attending a rehabilitation programme. Among controls, predictors of new symptoms or worsening with PCS development were an intercurrent secondary SARS-CoV-2 infection and educational status. At phase 2, persistent cases were less frequently never smokers, had higher values for BMI and body fat, and had lower educational status than controls. Fatigue/exhaustion, neurocognitive disturbance, chest symptoms/breathlessness and anxiety/depression/sleep problems remained the predominant symptom clusters, and exercise intolerance with post-exertional malaise for >14 h (PEM) and symptoms compatible with ME/CFS (according to Canadian consensus criteria) were reported by 35.6% and 11.6% of persistent cases, respectively. In adjusted analyses, significant differences between persistent cases and stable controls (at phase 2) were observed for neurocognitive test performances, scores for perceived stress and subjective cognitive disturbances, symptoms indicating dysautonomia, depression and anxiety, sleep quality, fatigue, and quality of life. In persistent cases, handgrip strength, maximal oxygen consumption, and ventilator efficiency were significantly reduced. However, there were no differences in measures of systolic and diastolic cardiac function, in the level of pro-BNP blood levels or other laboratory measurements (including complement activity, serological markers of EBV reactivation, inflammatory and coagulation markers, cortisol, ACTH and DHEA-S serum levels). Screening for viral persistence (based on PCR in stool samples and SARS-CoV-2 spike antigen levels in plasma in a subgroup of the cases) was negative. Sensitivity analyses (pre-existing illness/comorbidity, obesity, PEM, medical care of the index acute infection) revealed similar findings and showed that persistent cases with PEM reported more pain symptoms and had worse results in almost all tests. Conclusions: This nested population-based case-control study demonstrates that the majority of PCS cases do not recover in the second year of their illness, with patterns of reported symptoms remaining essentially similar, nonspecific and dominated by fatigue, exercise intolerance and cognitive complaints. We found objective signs of cognitive deficits and reduced exercise capacity likely to be unrelated to primary cardiac or pulmonary dysfunction in some of the cases, but there was no major pathology in laboratory investigations. A history of PEM >14 h which was associated with more severe symptoms as well as with more objective signs of disease may be a pragmatic means to stratify cases for disease severity.