Greenland is the largest island in the world, with a total area of about 836,000 square miles (2.17 million square kilometers) and a coastline of roughly 27,400 miles (44,087 kilometers). It is roughly three times the size of Madagascar, the next-largest island, and more than twice the size of New Guinea. That answer holds only if you accept the standard geographic convention that excludes continental landmasses from the island category. The convention is doing more work than most short explainers admit.
How Geographers Draw the Island Line
The core mechanism is a definitional filter applied before any measurement happens. Geographers classify landmasses into three broad tiers: continents, islands, and everything in between. The boundary between continent and island is not set by area, population, or geology alone. It is set by convention, and the convention is applied inconsistently across the world's landmasses.
Australia is the clearest example. It sits on its own tectonic plate, hosts a distinctive biota, and spans about 2.97 million square miles (7.69 million square kilometers). That is roughly 3.5 times Greenland's area. No geographer calls Australia an island. The reason is historical and taxonomic: Australia was classified as a continent before the island category was formalized, and the classification stuck. If you applied an area threshold alone, Australia would win by a wide margin.
Greenland itself sits on the North American tectonic plate, not its own plate. That is a strike against treating it as a mini-continent. Its status as an island rests on the combination of being smaller than Australia and being politically and cartographically treated as an island by major authorities. The United Nations Statistics Division and the CIA World Factbook both list Greenland as the world's largest island.
The mechanism, then, is not a measurement. It is a classification decision made upstream of the measurement. Change the classification rule and the answer changes.
Quantified Context: How the Top Islands Compare
The margin between Greenland and its nearest competitors is large, which is why the answer is stable under the standard convention. The table below compares the major contenders using total area and coastline length, with the caveat that coastline figures are notoriously sensitive to measurement method.
| Island | Total area (sq mi) | Total area (sq km) | Coastline (mi) | Notes |
|---|---|---|---|---|
| Greenland | ~836,000 | ~2,166,000 | ~27,400 | Largest under standard convention |
| New Guinea | ~303,000 | ~785,000 | ~8,500 | Divided between Indonesia and Papua New Guinea |
| Borneo | ~287,000 | ~743,000 | ~3,100 | Shared by Indonesia, Malaysia, Brunei |
| Madagascar | ~226,000 | ~587,000 | ~3,000 | Single sovereign state |
| Baffin Island | ~196,000 | ~507,000 | ~17,000 | Canada's largest island |
| Sumatra | ~182,000 | ~473,000 | ~3,000 | Entirely Indonesian |
| Honshu | ~87,000 | ~226,000 | ~6,800 | Japan's main island |
Greenland's lead over New Guinea is about 533,000 square miles, a gap larger than the entire area of South Africa. That is not a close call. The ranking among the top seven is also stable across data sources, even though absolute areas vary slightly depending on whether ice cover, tidal zones, and coastal inlets are included.
One caution on coastline: Greenland's figure of roughly 27,400 miles is measured at a specific map resolution. Coastlines are fractal, meaning measured length increases as the measurement unit shrinks. A finer-resolution measurement would yield a much larger number. The coastline ranking is therefore less reliable than the area ranking. Area is the better metric for this comparison because it is less sensitive to measurement scale.
Why Greenland's Ice Sheet Complicates the Count
About 81 percent of Greenland's surface is covered by an ice sheet, and the ice sheet is thick. At its center it reaches roughly 2 miles (3 kilometers) in depth. If you removed all the ice, the bedrock beneath would rebound isostatically over thousands of years, and the island's area would change. The contemporary area figure is therefore a snapshot of a glaciated landmass, not a fixed geometric fact.
This matters for the comparison because ice cover inflates the apparent land area relative to what a deglaciated Greenland would present. The bedrock surface is depressed below sea level in central Greenland due to the ice load. Some estimates suggest a fully rebounded Greenland would have a somewhat different coastline and area, though the change would not be large enough to unseat it from the top spot.
A counterintuitive detail: the ice sheet is also losing mass. Satellite gravimetry missions have documented sustained ice loss from Greenland since the early 2000s. The rate has varied year to year, with some years showing near-balance and others showing rapid loss. Over multi-decade timescales the trend is negative. This does not make Greenland smaller in a way that changes the ranking, but it does mean the "area" figure is a moving target at the margins.
Plain language: Greenland is still Greenland-sized, and the ice loss is a climate signal, not a ranking threat.
The Australia Problem and Other Misconceptions
The most common misconception is that Australia is an island and should therefore be the answer. It is not classified as one, and the classification has real consequences for how atlases, statistics agencies, and geographic databases are organized. Australia appears in continent tables, not island tables.
A second misconception is that Greenland's area figure includes only land. It includes the ice sheet, which sits on land, so the figure is defensible as a land area. It does not include the surrounding ocean or the territorial waters. Some sources quote a smaller "ice-free area" of roughly 158,000 square miles (410,000 square kilometers), which is a legitimate alternative metric but not the one used in the standard ranking.
A third misconception is that island size rankings are precise. They are not. Different sources use different coastlines, different treatment of inlets and offshore rocks, and different projections. The top spot is not in dispute, but the exact numbers for lower-ranked islands can vary by a few percent between sources.
Practical Actions: How to Verify the Ranking Yourself
If you need to check or cite the ranking, use a consistent method rather than mixing sources.
Pick one authoritative dataset and stick to it. The CIA World Factbook and the UN Statistics Division both publish island and country area figures. Use one for all comparisons. Confirm the classification rule before comparing. Check whether the source treats Australia as a continent and whether it includes ice-covered area. Compare area, not coastline. Coastline length is scale-dependent and produces unstable rankings. Note the measurement date. Ice cover, coastal change, and political boundaries shift slowly but do change. Distinguish total area from ice-free area when citing Greenland, and state which figure you are using. Do not compare island figures to continental figures without flagging the category difference.
Red-flag warnings when reading secondary sources:
A source that lists Australia as the largest island is using a nonstandard classification. A source that gives Greenland an area below 800,000 square miles is likely using an ice-free or otherwise restricted figure. A source that ranks islands by coastline length without stating resolution is not reproducible. A source that mixes square miles and square kilometers without conversion invites arithmetic errors.
Limitations and the Definitional Frontier
The largest-island question has a clean answer only because the continent/island boundary is treated as settled. It is not settled in any physical sense. There is no agreed area threshold that separates a continent from an island. Australia is a continent because it has been called one, and Greenland is an island because it has been called one. The distinction is taxonomic, not empirical.
That means the answer to "which is the largest island in the world" depends on a convention you accept before you measure anything. Under the standard convention used by major geographic authorities, Greenland is the answer by a wide margin. Under an area-only rule that ignores the continent category, Australia would win. Under a rule that excludes ice-covered land, the ranking would shift slightly but Greenland would likely remain first.
The honest formulation is this: Greenland is the largest island under the classification system in general use, and the system is a human decision rather than a fact of nature. That distinction does not undermine the answer. It explains why the answer is stable and why it is worth stating with the convention attached.