Indonesia has the most followers of Islam relative to the "over 87% of the population" threshold: roughly 87–88% of its residents are Muslim, and because the country holds about 240 million Muslims, no other nation combines that share with that absolute size. Afghanistan, Somalia, and Morocco post higher percentages, but they are far smaller. The precise ranking depends on which dataset and which year you use, and on whether you count adults, citizens, or residents.
That answer sounds simple. The number hiding behind it is not. "Over 87% of the population" is a threshold that shifts with every census, every migration wave, and every change in how a state counts religion. This article explains the mechanism that produces these shares, quantifies the global picture, and shows where the rankings break down.
How a population share is actually produced
A national Muslim share is a ratio, not a count. The numerator is the number of people a survey or census classifies as Muslim. The denominator is the total population the same instrument covers. Change either one and the percentage moves.
Three mechanisms drive that ratio. First is natural growth: fertility and mortality rates differ by religious group, so shares drift over a generation. Second is migration, which can move a share sharply in small countries. Third is definitional: some censuses ask religion directly, others infer it from ethnicity or nationality, and some refuse to ask at all.
The definitional layer is where most confusion starts. A country that counts only citizens will report a different Muslim share than one counting all residents, because migrant workers are often excluded from citizenship but included in population totals. A country that asks about religion on a mandatory census produces cleaner data than one relying on household surveys with nonresponse.
Plain language: the percentage is a photograph of two moving numbers, taken with a camera whose lens differs by country. Comparing shares across borders is comparing photographs taken with different cameras.
The quantified global picture
The most-cited global reference is the Pew Research Center's demographic work, which estimated about 1.9 billion Muslims worldwide around 2020, roughly a quarter of humanity. The largest Muslim populations by absolute count sit in Indonesia, Pakistan, India, Bangladesh, and Nigeria. Only some of those clear the 87% share threshold.
The countries that reliably clear 87% share are mostly smaller or middle-sized states. Afghanistan, Somalia, and Morocco are commonly estimated above 99%. Iran, Algeria, Tunisia, and Yemen typically land in the high 90s. Indonesia sits just under the threshold depending on the source, which is why the "over 87%" framing is fragile.
A representative comparison shows how share and size diverge:
| Country | Approximate Muslim share | Approximate Muslim population | Why it matters |
|---|---|---|---|
| Indonesia | 87–88% | ~240 million | Largest absolute count, share near threshold |
| Pakistan | 96–97% | ~240 million | Huge count and high share |
| India | 14–15% | ~200 million | Large count, low share |
| Nigeria | 50–53% | ~110 million | Split roughly evenly |
| Afghanistan | ~99.7% | ~40 million | Highest share, small count |
| Somalia | ~99.8% | ~18 million | Highest share, tiny count |
| Morocco | ~99% | ~37 million | High share, mid-size count |
The pattern is clear. High shares cluster in North Africa, the Horn of Africa, and West Asia. High absolute counts cluster in South and Southeast Asia. Only a handful of countries sit in both zones at once, and Indonesia and Pakistan are the strongest candidates.
Where the threshold breaks down
The 87% line is not a natural boundary. It is an arbitrary cut, and several data problems make it slippery.
Census timing is the first problem. National censuses run on five- to ten-year cycles, and some countries have postponed or canceled them. A share reported for 2010 may be the only official figure available in 2024. Presenting it as current overstates precision.
Nonresponse is the second problem. Survey refusals are not random. In some countries, religious minorities decline to answer, which inflates the reported majority share. In others, the state discourages reporting minority religion at all.
The third problem is the citizenship filter. Gulf states count large expatriate populations who are not citizens and are often not counted by religion in official statistics. That changes the denominator in ways that are rarely disclosed.
Fourth is the definition of "Muslim" itself. Some instruments count cultural or nominal affiliation, others count only self-identified practice. The gap between the two can reach several percentage points in countries with strong secular traditions.
Here is the counterintuitive detail: a country can have a higher Muslim share than Indonesia and still be irrelevant to the keyword question, because the question implicitly asks about scale as well as share. Afghanistan's share is nearly perfect and its Muslim population is roughly one-sixth of Indonesia's.
Common misconceptions about Muslim-majority rankings
Several ideas circulate that the data does not support.
Misconception one: the largest Muslim population is in the Middle East. In fact, most Muslims live in Asia, and the single largest national count is in Southeast Asia. Misconception two: a high percentage means a large absolute number. Small states can hit 99% with populations under 20 million. Misconception three: shares are static. Fertility and migration shift them measurably within a single generation. Misconception four: every country counts religion the same way. Census instruments differ enough that cross-border ranking is approximate. Misconception five: the 87% figure is an official statistic. It is a threshold chosen for a query, not a published benchmark. Warning: any single-number ranking that does not name its source year and definition should be treated as a rough estimate.
The most persistent error is treating a percentage as a count. A 99% share tells you almost nothing about how many people that represents.
Practical steps for reading these numbers correctly
If you need a defensible answer rather than a headline, work through these steps in order.
First, name the metric. Decide whether you want share, absolute count, or both. The keyword question blends them, so state which one you are answering.
Second, name the year. Use the most recent census or the Pew estimate and label it explicitly. Do not present a decade-old figure as current.
Third, check the denominator. Confirm whether the source counts all residents, citizens only, or adults only. The gap can exceed five percentage points in countries with large migrant workforces.
Fourth, set a minimum population. If you require both high share and large count, a threshold of 30 million residents filters out the small states and leaves a short list.
Fifth, reconcile the thresholds. A country at 87.1% passes the literal cut but sits within the margin of error of a 2010 census. Treat shares within two percentage points of a threshold as unresolved.
Plain language: pick your definition, pick your year, and pick your minimum size before you rank anything. Otherwise the ranking is an opinion wearing a number.
A time course for how shares evolve
Religious shares are not fixed, and the direction of change is not uniform. The table below shows the general pattern rather than country-specific measurements.
| Period | Dominant mechanism | Effect on majority shares |
|---|---|---|
| Pre-1950 | High fertility, limited migration | Shares stable over decades |
| 1950–1980 | Postcolonial population growth | Large absolute increases, shares roughly steady |
| 1980–2000 | Labor migration to Gulf states | Denominator shifts in host countries |
| 2000–2020 | Urbanization, falling fertility | Share growth slows in several majority countries |
| 2020–2050 (projected) | Diverging fertility by region | Some shares rise, others plateau or decline |
The projected row is the weakest. Long-range religious projections carry wide uncertainty bands because they compound assumptions about fertility, mortality, and migration. Treat any 2050 share as a scenario, not a forecast.
This is the professional caveat worth repeating: demographic projections are conditional statements. They say what happens if current rates hold, and current rates rarely hold.
Limits of what the data can support
Several limits bound any answer to this question.
Data quality varies enormously by country. Some national statistics offices publish detailed religion tables. Others publish none, leaving international compilers to model from surveys. A ranking that mixes the two is not measuring the same thing twice.
Religion is self-reported in most instruments, which means it captures identity rather than practice. A country can be 99% Muslim by identity and far lower by regular attendance. The keyword asks about followers, and "follower" is not a category any census measures directly.
The threshold itself is arbitrary. Any cut at 87% will place some countries just above and some just below on the strength of a rounding choice. The honest answer names Indonesia and Pakistan as the two countries that combine a share near or above the line with a population above 200 million, and notes that the exact ordering depends on the source.
Finally, no single article can substitute for the underlying datasets. The Pew Research Center's religion-and-public-life demographic reports and national census publications from each country are the primary sources. Where those disagree, the disagreement is the finding.
This article is general educational information about demographic measurement, not a legal, immigration, or policy opinion. For decisions that depend on population data, consult the primary statistical source and a qualified local professional.