I. Executive summary The Austrian Science Fund (FWF), which is Austria's main funding organisation for basic research, encourages and helps all project leaders and project staff members to make their peer-reviewed research results freely available on the Internet. Since 1 January 2016, open access has been mandatory for all peer-reviewed publications listed in final reports on FWF-funded projects. Any exceptions must be clearly indicated and justified. For projects that started after 1 January 2015, open access has been compulsory for all peer-reviewed publications. All principal investigators in FWF-funded projects are obliged to submit a final report within three months of completing their projects. The publications and other data mentioned in those reports are archived and analysed by the FWF. This report examines the state of compliance with open access requirements on the basis of final project reports submitted in the year 2016. Main findings: A total of 6,994 publications were listed in the final project reports submitted in 2016. Of those publications, 5,146 were conclusively identified as peer-reviewed. Regarding compliance with the FWF's Open Access Policy, the report shows that 92% of the peer-reviewed publications that arose from FWF projects in 2016 are openly accessible (2015: 83%). The most frequently chosen option is hybrid open access (40%). Since last year's report, the use of green open access has increased by 3%, other open access by 5%, and gold open access by 2%. The majority of peer-reviewed publications submitted are journal articles, 93% of which are openly accessible. The lowest rate of compliance with the FWF's Open Access Policy can be found in editions, contributions to edited volumes and monographs (58%). Although open access is not compulsory, 48% of non peer-reviewed publications are freely available through the internet. The absolute number of publications listed in final reports in 2016 includes 321 publications that are mentioned several times in different projects; however, those repetitions do not ultimately affect the relative share of open access publications (92%). II. Bias All publications were entered in the FWF’s database manually and with the utmost care when the final project reports were received. Nevertheless, it is not possible to rule out minor errors entirely. Moreover, some systematic problems need to be solved in the future; the FWF is currently preparing to implement a research information system in which FWF-funded researchers will be able to enter and update their research data on an ongoing basis. Due to the amount of data involved, it was not possible to check the current open access status of all publications twice. The last check was performed in January 2017. In some cases, the embargo period may have expired after submission, and publications labelled "Other open access" in the FWF’s database may now be accessible through green open access. Whenever the status "peer-reviewed" could not be clearly identified according to FWF guidelines, the publications were classified as non peer-reviewed.

We are grateful to Katharina Rieck for support with proofreading. Any remaining errors are our own.