How to explore: You can hover over the charts for details; zoom and toggle categories on the map. Counts refer to recorded incidents, not population-adjusted risk.
The shark attack history in Australia has two distinct trends. While the number of recorded attacks has risen in recent times, this is not the case with the proportion of recorded incidents resulting in fatalities.
Between 1900 and 1949, 105 of 308 recorded incidents were fatal (34.1%). Between 2000 and 2019, 35 of 398 incidents were fatal (8.8%). These periods differ in length, so the interactive chart uses annual counts and five-year averages rather than comparing their raw totals directly.
The record also reflects changes in reporting and completeness over time.
The Recorded incidents are not evenly distributed across the Australian coastline. New South Wales leads with the most number of attacks with 469 recorded incidents, followed by Queensland (377, after standardising the state abbreviation) and Western Australia (242). Together they account for about 84% of the 1,293 incidents with known outcomes in this analysis.
But the state with the most recorded incidents is not the state with the most fatalities: Queensland has 102, compared with 82 in New South Wales. South Australia has a fatal share of 29.4% among its 85 recorded incidents.
The number of recorded swimming cases is 473, which includes 159 fatalities (33.6%). In comparison, boarding comprises 280 cases with 26 fatalities (9.3%). The fatality rate is 20.6% for snorkelling (27 out of 131) and 22.0% for diving (20 out of 91).
This is only a percentage of cases that were documented and not the probability of an incident happening in each case. We do not know how many people actually indulge in the above activities and spend time in the water. Cases below 10 have been left out of the above table for clarity.
Comparing just species could hide essential differences. The fatality rate for white shark is 42.7% when swimming (41 of 96), in comparison to 11.8% when boarding (21 of 178). For tiger shark incidents that involve swimming, out of 100 incidents 54 incidents resulted in death, or 54.0% of the total.
There are some combinations with few data points. For example, incidents of white sharks snorkeling, there are 6 fatalities in 11 cases (54.5%). This percentage cannot be considered a reliable rating of risk. The heatmap presents the top 6 species and the top 7 activities based on how often they are mentioned in the data set; the combinations with fewer than 8 occurrences have been left blank.
In the 1920s, 32 of 81 recorded incidents were fatal (39.5%). In the 2000s and 2010s, the proportions were 8.5% (14 of 165) and 9.0% (21 of 233), respectively.
This trend does not have to do with a continual decrease either, because for the period between 2020 and 2026, 24 out of 164 instances resulted in fatalities (14.6%). This decade has the greater percentage in terms of fatalities compared to the previous two decades, but the 95% Wilson intervals overlap (for the 2020s: 10.0–20.9%; for the 2010s: 6.0–13.4%).
These figures clearly demonstrate that shark encounters are variable based on time, geography, human action, and even species. While numbers are useful, they do not necessarily denote an individual risk level. There is no consistent denominator through time of numbers of people on beaches, how long people spend in the water and what activities people engage in. It is likely that older occurrences have been lost, there are definitional differences, and changes in practice for reporting. As such, the percentage of fatal attacks from records is not equal to risk from entering the water for humans and encountering sharks.
It is not the case that sharks have gotten less dangerous and that Australia is unequivocally safer when it comes to shark attacks. What the data demonstrate is that more recorded shark encounters do not necessarily mean that a greater proportion of those encounters result in fatalities..
Data source: Australian Shark Incident Database (ASID), Taronga Conservation Society Australia. Official dataset page: https://www.taronga.org.au/wildlife-and-conservation/the-institute/australian-shark-incident-database.
Data preparation: Cleaning was done for the Australian Shark Incident Database, so that standardized category coding and treatment of missing data were possible. In order to highlight previous trends, five-year moving averages have been used, and in order to make comparisons clearer, minimum sample size cut-offs have been applied..
Acknowledgement: Data sourced from the Australian Shark Incident Database, Taronga Conservation Society Australia, in partnership with the organisations credited on its official data page. Interpretations and visualisations are the author’s own and are not endorsements by the data provider.
Generative AI declaration: Generative AI tools were used for R code debugging and paraphrasing informal phrasing into an academic tone. The core story, topic, analytical direction, chart choices, and interpretations are entirely the author’s original work. The author reviewed all output against the original ASID data, made necessary revisions, and remains fully responsible for the accuracy, interpretation, and academic integrity of the submitted work.
Taronga Conservation Society Australia. (n.d.). Australian Shark Incident Database. https://www.taronga.org.au/wildlife-and-conservation/the-institute/australian-shark-incident-database OpenAI. (2026). ChatGPT [Large language model]. https://chatgpt.com/