Australia's remote health sector has chronic understaffing issues and serves an isolated, culturally diverse population with a high burden of disease. Workplace health and safety (WHS) impacts the wellbeing and sustainability of the remote health workforce. Additionally, poor WHS contributes to burnout, high turnover of staff and reduced quality of care. The issue of poor WHS in Australian very remote primary health clinics was highlighted by the murder of remote area nurse (RAN) Gayle Woodford in 2016. Following her death, a national call for change led by peak bodies and Gayle's family resulted in the development of many WHS recommendations and strategies for the remote health sector. However, it is unclear whether they have been implemented. The aim of this study is to identify which WHS recommendations have been implemented, from the perspective of RANs. A cross-sectional online survey of 173 RANs was conducted during December 2020 and January 2021. The survey was open to all RANs who had worked in a very remote (MM 7 of the Modified Monash (MM) Model) primary health clinic in Australia more recently than January 2019. A convenience sampling approach was used. The survey tool was developed by the project team using a combination of validated tools and remote-specific workplace safety recommendations. Broad recommendations, such as having a safe clinic building, safe staff accommodation, local orientation, and 'never alone' policy, were broken down into specific safety criteria. These criteria were used to generate workplace safety scores to quantify how well each recommendation had been met, and clustered into the following domains: preparation of staff, safe work environment and safe work practices. Descriptive statistics were used and the safety scores between different states and territories were also compared. Overall, the average national workplace safety score was 53% (standard deviation (SD) 19.8%) of recommendations met in participants' most recent workplace, with median 38.5% (interquartile range (IQR) 15.4-61.5%) of staff preparation recommendations, median 59.4% (IQR 43.8-78.1%) of safe work environment recommendations, and median 50.0% (IQR 30.0-66.7%) of safe work practices recommendations met. Within domains, some recommendations had greater uptake than others, and the safety scores of different states/territories also varied. Significant variation was found between the Northern Territory (57.5%, SD 18.7%) and Queensland (41.7%, SD 16.7%) (p<0.01), and between South Australia (74.5%, IQR 35.9%) and Queensland (p<0.05). Last, many RANs were still expected to attend after-hours call-outs on their own, with only 64.1% (n=107/167) of participants reporting a 'never alone' policy or process in their workplace. The evidence from this study revealed that some recommended safety strategies had been implemented, but significant gaps remained around staff preparation, fatigue management and infrastructure safety. Ongoing poor WHS likely contributes to the persistently high turnover of RANs, negatively affecting the quality and continuity of health care in remote communities. Variation in safety scores between regions warns of a fragmentation of approaches to WHS within the remote health sector, despite the almost identical WHS legislation in different states/territories. These gaps highlight the need to establish and enforce a national minimum standard of workplace safety in the remote health sector.