Mid-March 2020, due to the worldwide spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and first reports of severe and fatal cases of the increasingly fast-spreading COVID-19 disease, Germany closed all of its schools as part of its first national lockdown, aimed to curb the pandemic spread of COVID-19. This first, 9 weeks enduring lockdown caused various social, economic and health problems1 and aggravated societal inequalities,2 which led to increasing pressure for the recommencement of regular in-school teaching. In order to assure the safety of school re-openings in Saxony, it was decided to scientifically accompany this throughout 2020 to monitor the potentially non-symptomatic spread of SARS-CoV-2 in schools. We longitudinally screened schools in four different districts in Saxony (Germany) for non-symptomatic active SARS-CoV-2 infections as well as seroprevalence in students and staff in three consecutive rounds (I-III) of examinations throughout 2020, cf. Figure 1. The study was approved by the Ethics Committee of the Medical Faculty of the University of Leipzig (Reg. No. 260/20-ek). All procedures conform to the declaration of Helsinki. The present study is registered in the German Clinical Trials Register (ID: DRKS00022032) as well as with the Robert Koch Institute as seroepidemiological study. In total, we tested 3,053 non-symptomatic participants in 18 schools onsite. About 60% of the participants were females, and this imbalance was mainly caused by predominantly female staff in primary (ca. 80%) and secondary (ca. 70%) schools. In sum, we analysed 7,472 oropharyngeal swabs to screen for acute infections using real-time PCR and 6,715 blood serum samples to test for SARS-CoV-2 IgG antibodies. Online questionnaires taken by about 70% of the participants provided additional information, for example regarding previous symptoms or contact to identified COVID-19 cases. During examination round (I), we found none, and in (II), we found one active infection with SARS-CoV-2. In contrast, round (III) revealed 26 SARS-CoV-2 positives out of 2,449 swab samples. As such, our observed incidence increased overall from 0·0% in spring to 0·04% in summer and up to 1·1% in autumn 2020, expectedly the relative detected infections increased with the varying regional 7-day incidence/100,000 inhabitants (7-di). The fact that in regions with the highest local incidence (300–400 7-di) disproportionately high numbers (22/717) of symptomless cases in schools were found compared to lower incidence regions (4/1748 in 100–150 7-di and 1/2438 in 1–10 7-di) could indicate that schools follow the overall pandemic development rather than driving it. Analysing the differences in infection rates between the tested groups in round (III), we found higher incidences in secondary schools (students 1·9%, staff 1·0%) than in primary schools (students 0·5%, staff 0·3%). In both school forms, students showed a higher rate of non-symptomatic infections than the respective staff, as might be expected since adolescents, and particularly, children are less symptomatic than adults3 and therefore more likely to carry undetected infections. Particularly striking is the difference in rate between 7th grade students (2·6% = 13/498) and 11th grade (1·0% = 4/401). This could not be shown to be significant (p = 0·08), but we note that the relative risk for infection with SARS-CoV-2 was 2·6-fold higher for grade 7 compared with grade 11, where only the latter were required to wear masks inside the classroom. Other studies found similar results for mandatory measures.4 Answers to the questionnaire revealed that 48% (10/21) of the currently infected participants reported symptoms within the previous 10 days compared to only 11·9% (202/1700) of the non-infected (p = 7·3*10−7). Symptoms of the respective household members also showed significant differences (p = 0·01). The seroprevalence was initially 0·6% in spring (I), remained constant over the summer (0·6% in (II)) and increased in the second pandemic wave to 1·4% in (III), so it was expectedly following the trend of the overall incidence. Our detected seroprevalence thereby exceeds officially reported COVID-19 cases by a factor of about five in rounds (I) and (II), similar to other studies.5 With this factor the seroprevalence increase we found in (III) trails the increase in officially reported cases by three to four weeks, which is longer than the expected one to two weeks for antibody development. Again, this is consistent with schools rather following the surrounding 7-day incidence. In conclusion, we found that school openings at low incidence are safe, but insufficient safety measures, especially at high incidence, are associated with more infections in schools. Nonetheless, results indicate that schools follow the surrounding pandemic incidence rather than driving it, even at elevated regional incidence. Rigorous testing of students and staff as well as mandatory mask-wearing in all age groups appear to be indispensable for safe, continuously open schools during elevated regional incidence. We thank the schools and all participants, their families, the LIFE Child-staff and supporting paediatricians, as well as M. Vogel for obtaining project funding from the Free State of Saxony. All authors declare no competing interests. This study was supported by the Free State of Saxony, Germany.