Background: Pancreatic cancer has one of the lowest survival rates of all cancer types, with an incidence to mortality ratio approaching one. People with pancreatic cancer experience a rapid decline in health characterized by pain, nausea, fatigue and weight loss. For most people, the disease is detected at an advanced stage and the focus of treatment is palliative. In Victoria, Australia, knowledge regarding patterns of care for people with pancreatic cancer is out-of-date, but central to quality improvement initiatives targeting unwarranted variations in care and improvement in supports that are consistent with patient preferences. Aim: Our aim was to compare care received by patients with pancreatic cancer with a consensus-based standard representing optimal care to identify deviations from best practice and highlight processes that may improve the quality and safety of care provided. Methods: Eligible patients included those with pancreatic cancer, first treated in 2015, at one of three tertiary hospitals in Victoria, Australia. Once identified, dates and details of events indicated by the optimal care pathway were extracted from the medical record of each patient. Data were summarized using descriptive statistics and process maps: a visualization method that illuminates gaps, duplication, deviations from best practice and processes that may be amenable to improvement. Results: Thirty-two of 165 care pathways have been mapped to date. The nature and timing of care received appears highly variable. Only nine of 32 patients (28%) received all of their cancer care at a single institution; the remainder (n=23, 72%) received care in multiple tertiary and community facilities. Apart from four (13%) emergency presentations, referrals for specialist care came from general/primary practitioners (n=26, 81%). The timeframe for general/primary practitioner investigations ranged from one to 57 days. Once referred to a tertiary setting, most patients (n=23, 72%) were discussed at a multidisciplinary team meeting and received standard therapies. Only four had resectable disease. Nineteen patients (60%) had documented referrals to hospital- or community-based palliative care services. Where observed, deviations from the consensus-based standard tended to be related to the difficult nature of diagnosing pancreatic cancer, and determining appropriate care for patients with an advanced cancer with nonspecific symptoms. Conclusion: Process mapping provided a useful and efficient means of comparing care received with a consensus-based standard; however, the assessment of adherence to optimal timeframes and specific care events was complicated by missing data. Implications for quality improvement activities will be considered in the context of study limitations. We will also emphasize the importance of engaging patients and carers in setting improvement priorities.