newy.com.au – A coach driver who was likely impaired by tramadol was travelling too fast to negotiate a tightening right-hand curve within the Branxton Interchange, an exhaustive investigation into the Greta coach crash has found. The coach’s windows then failed to contain passengers, with all 10 deaths associated with complete or partial ejection.
The Office of Transport Safety Investigations’ 287-page final report (PDF), released on Friday, identified excessive speed and likely tramadol impairment as factors contributing to the rollover, and the failure to contain passengers and incomplete seatbelt use as factors contributing to the severity of its consequences. It also identified wider shortcomings in commercial driver health oversight, cross-employer performance monitoring, road design guidance and vehicle safety standards. Ten passengers died, the other 25 were hospitalised with injuries of varying severity and the report acknowledged the profound psychological harm suffered by survivors, families, first responders and the wider community.
Coach 4666MO, a 57-seat Volvo operated by Linq Buslines, left a Lovedale winery at 11.17pm on 11 June 2023 with 35 passengers returning to Singleton. It travelled through Greta and approached the grade-separated elliptical roundabout from Wine Country Drive, intending to turn right towards the Hunter Expressway’s westbound on-ramp. The driver was familiar with the interchange, had passed through it twice earlier that day and later said they knew it was a sharp corner.
Wine Country Drive and the interchange were within an 80km/h speed zone, but OTSI’s modelling found much lower speeds were required through the roundabout’s compound curves. On-board telematics recorded the coach at 52.6km/h about 40 metres before it entered the roundabout and at 56.48km/h on the overpass about 130 metres before the rollover. The exact speed at the point of rollover was not recorded. OTSI’s reconstruction found the coach was probably travelling between 47km/h and 57km/h as the right-hand curve tightened, above a theoretical rollover threshold of 47km/h at the critical point. The model put the safer speed at that location, allowing a margin against rollover, at 37km/h.
The coach understeered and began to yaw across the lane markings as its weight shifted onto the left-side tyres. Its right-side wheels lifted and it rolled onto its left side, struck the road and barrier kerb, then slid across the cycle path into a guardrail. The guardrail rotated the front of the coach back towards the road and probably prevented it continuing into vegetation and a gully.
OTSI found no indication that fatigue, microsleep, distraction, driver inexperience, poor visibility, weather, fog, a mechanical fault or the vehicle travelling behind the coach contributed to the crash. Investigators also found no evidence of heavy braking before the rollover. Although skid resistance was below an investigatory level in parts of the roundabout, it did not contribute.
An occupant of the following vehicle called Triple Zero at 11.33pm, followed by a passenger on the coach a minute later. Crash detection on three passengers’ Apple devices also alerted emergency services. Police officers who had been at the nearby Greta service station arrived about 11.38pm and the first ambulance arrived about two minutes later. Occupants of the following vehicle helped passengers remove the coach’s two roof escape hatches, while police assisted people through the hatches and helped remove the windscreen to gain access.
Post-crash testing found no alcohol in the driver’s system. A blood sample taken about two hours after the rollover contained 0.62mg/L of tramadol and more than 5mg/L of paracetamol. A specialist pharmacologist found the tramadol concentration was inconsistent with the prescribed amounts and was likely to have impaired the driver’s sensory, cognitive and motor functions, although the degree of impairment and exact amount taken could not be determined. During the criminal proceedings, the driver said they had taken 350mg of tramadol that day, including a 150mg slow-release capsule about 10pm, bringing the stated daily total to 100mg above the prescribed maximum.
“My report found the driver should not have been behind the wheel that night, given a history of likely misuse of the opioid prescription painkiller tramadol, including likely impairment at the time of the crash,” [OTSI Chief Investigator Jim Modrouvanos](https://www.otsi.nsw.gov.au/news-and-media/media-release-rollover-coach-4666mo-greta-11-june-2023) said.
The driver had been prescribed tramadol for chronic pain since at least 2011. Medical and pharmacy records led OTSI to find likely opioid dependence and a history of medication misuse. Two pain specialists raised concerns in 2022 and recommended further steps, including an addiction assessment and review of whether the medication was compatible with driving a public passenger vehicle, but the addiction assessment was not completed.
The same general practitioner completed the driver’s commercial fitness-to-drive assessments in 2016, 2019, 2021 and 2022. Tramadol was declared in 2016, when the prescribed dose was reported as 50mg a day and compatible with driving, but it was not included in the 2019, 2021 or October 2022 assessments. By the final assessment, prescriptions were higher and specialist concerns had been raised. OTSI could not determine why the assessments varied. It could not conclude that complete assessments would necessarily have disqualified the driver in the earlier years, but found disclosure of the substance-misuse concerns in 2022 would probably have removed the driver from commercial duties for at least a period.
The driver’s former employer, CDC NSW, removed the driver from driving duties in June 2022 after receiving the first specialist report. An independent medical assessment was scheduled for 19 October, but the driver resigned that day before it took place. CDC told OTSI it did not notify Transport for NSW because it had not received a definitive opinion about fitness to drive. Separately, privacy restrictions prevented the company sharing sensitive medical information with later employers without consent.
The driver then worked briefly for Sid Fogg’s Coaches, resigning after management raised concerns about errors and omissions in the National Driver Work Diary and driving when instructed only to observe.
When Linq hired the driver on 21 November 2022, it confirmed there was a valid heavy rigid licence, Bus Driver Authority and medical assessment. Its reference check was a short text-message exchange that did not cover the driver’s health, reasons for leaving CDC or earlier performance history.
CDC had received 27 customer complaints and one compliment about the driver between 2016 and 2022, with 12 complaints substantiated and three partially substantiated. The substantiated matters included erratic driving, pulling out in front of vehicles, near misses with cyclists, harsh acceleration and braking, tailgating and travelling too fast for the conditions. CDC also recorded 11 minor traffic incidents and issued six formal warnings in incidents where it found the driver at fault. None of that history was known to Linq.
Linq had recorded three earlier overspeed events involving the driver, two of which remained open at the time of the crash. A complaint received after the rollover about a school charter on 8 and 9 June was supported by telematics showing four occasions above 100km/h. The report noted that the coach was speed-limited to 100km/h but could exceed that speed while travelling downhill. OTSI found Linq’s system lacked clear thresholds for responding consistently to speeding, harsh braking, harsh acceleration and fatigue breaches, while NSW had no central system allowing successive operators to see a bus driver’s full complaint, incident, infringement or cross-employer performance history.
Routine industry drug screening was also incapable of detecting tramadol. The industry-standard urine tests were designed to detect specified drug classes, including naturally derived opiates such as morphine and codeine, but not the synthetic opioid. A Linq test in February 2023 returned a negative result even though the driver listed tramadol on the testing company’s paperwork. OTSI said the declaration did not appear to have been identified by Linq management and found no record of further inquiries. The report stressed that the test worked within its intended scope and was not designed to identify every potentially impairing medicine or establish impairment.
The investigation found the coach’s structural frame performed well and preserved the survival space with minimal intrusion. The windows did not provide the same protection. All windows on the coach’s left side broke, 14 passengers were fully ejected and three were partially ejected. Nine deaths were associated with complete ejection and one with partial ejection. Many of the 12 passengers with significant physical injuries were exposed to the road, kerb, cycle path or guardrail through the broken windows.
OTSI determined that six of the 35 passengers were wearing the available lap-sash seatbelts. Four of those passengers had minor injuries and two had significant injuries. One restrained passenger beside a window was partially ejected after the upper body slipped out of the shoulder sash, while the lap section prevented complete ejection. The report found the Australian glazing standard did not provide sufficient occupant containment in rollovers and identified further risks from typical seatbelt geometry, fixed shoulder-belt heights and variable windowsill heights.
The passengers were not given the pre-trip safety briefing required by Linq policy. Although the coach had at least one window sign and a front “FASTEN SEATBELT” sign, OTSI found neither Transport for NSW nor the Australian Design Rules specified minimum content, number or placement of signs. Wearing a fitted seatbelt was legally required, but Revenue NSW data reviewed by OTSI showed only one passenger aged over 16 was fined for failing to wear one on a coach between 1 July 2022 and 30 June 2025.
OTSI did not find the Branxton roundabout was a direct cause of the crash, and its design complied with the standards and guidelines that applied at the time. However, the report found its elliptical shape, compound curves, grade and tightening radius created risks that could have been better managed. An early design verifier warned that Branxton’s elliptical layout could produce inconsistent speeds and sideswipe crashes, while Kurri Kurri’s changing-radius design could increase sideswipe and truck-rollover risk. No related changes were made because the designs met the requirements and the roundabouts were treated as low-speed environments.
The report said additional geometric controls, warning signs and road markings could reduce speeds and heavy-vehicle rollover risk at the Branxton, Kurri Kurri and Buchanan interchanges. Transport for NSW advised it had developed a proposal for improved advance warning and more accurate directional signs at all three locations, but the proposal was still awaiting endorsement in May 2026.
OTSI made 34 recommendations to government agencies, regulators, operators and other organisations, along with seven broader recommendations to the bus and coach industry. The directed recommendations cover commercial-driver medication management and declaration requirements, health assessments, a central driver performance system, seatbelt education and enforcement, anti-ejection glazing, improved restraints and rollover exits, heavy-vehicle stability modelling, event recorders and automatic crash notification.
All 34 directed recommendations were accepted or accepted in principle, although the responses show that many measures remain subject to further review, feasibility work, funding or future rule-making. Three accepted recommendations directed to Linq were later marked suspended because the company no longer operates, although the report records safety briefing, speed management and risk register changes made after the crash. The National Transport Commission will consider the medical recommendations in its next Assessing Fitness to Drive review, due in 2027, while Austroads expects updated roundabout design guidance in mid-2028. The Commonwealth has accepted a recommendation to investigate anti-ejection glazing as part of a future Australian Design Rules work plan.
The NSW Government said it had accepted or accepted in principle all 21 recommendations directed to Transport for NSW and would consult the bus industry on what it described as the biggest reform of the state’s Bus Safety Regulatory Framework in 20 years. Proposed legislation would create new safety duties for operators, company officers, those who procure bus services, bus safety workers and drivers, backed by stronger compliance powers and a tiered system of offences and penalties. The government also intends to establish a new regulator, in consultation with the industry, to administer the framework and oversee compliance.
Transport Minister John Graham said the report provided a path to stronger safety across the bus network. “We can’t undo the immense loss and trauma caused by this crash, but we learn and change from it – and we are doing that,” he said.
The government said 33 of the NSW Bus Industry Taskforce’s 58 recommendations had been implemented, including all five recommendations in a safety report prepared in direct response to the crash. Measures have included updated Safety Management System Guidelines and Fit and Proper Person Policy, a technology initiative, revised annual self-assessments for operators, fitting seatbelts to more rural, regional and outer-metropolitan buses and the “Buckle Up on the Bus” campaign. Work on the remaining recommendations is continuing.
Roads and Regional Transport Minister Jenny Aitchison said the tragedy continued to affect families and communities, particularly in NSW and the Hunter. “That’s why safety must always remain at the centre of every decision we make across our road and transport network. To ensure that other families do not have to share this trauma and loss,” she said.
“My 41 recommendations are intended to deliver actions that improve safety systems and prevent future similar crashes,” Mr Modrouvanos said.
OTSI said its no-blame investigation was independent of police, Transport for NSW and other government agencies, and was not intended to assist court proceedings or determine liability. Separately, the driver pleaded guilty to 10 counts of dangerous driving occasioning death and other offences, received a 32-year prison sentence with a 24-year non-parole period and later had a sentence appeal dismissed.
