Getting a Read on the Number
The straightforward answer is that no one actually knows, and anyone giving you a single decimal-point figure is selling something. The best available data comes from large-scale survey programs like Gallup and academic work from the Williams Institute, and even those disagree with each other depending on how they phrase questions and which countries they cover. By conservative self-identification numbers — people who actively label themselves as lesbian, gay, or bisexual on a survey — the global figure lands somewhere between 3% and 5% of the total population. That puts it roughly in the range of 250 to 400 million people worldwide. If you broaden the definition to include anyone who has had same-sex attraction or experience at some point in their life, even if they don't identify as gay or bisexual, the numbers climb considerably, and different studies produce wildly different results depending on methodology. The problem is structural. In a lot of countries, openly identifying as gay carries real social or legal risk. You're not going to get honest answers from national censuses in places where homosexuality is criminalized, and even in countries where it's accepted, survey fatigue and social desirability bias skew results. I worked on a project a few years ago where we tried to model sexual orientation prevalence across Southeast Asia using census data and health survey fragments, and the variance was absurd. A single country's estimate could swing by four percentage points depending on whether the question was "have you ever had sex with another man" versus "do you identify as gay." That's not a measurement error, that's the difference between behavior and identity, and they don't always line up.
Gallup's 2023 data for the United States shows about 7.6% of adults identifying as LGBTQ+, with Gen Z hovering around 20%. That generational gap tells you everything you need to know about why global estimates are so messy. Younger cohorts are far more willing to use non-heterosexual labels, which means a snapshot from 2025 will look dramatically different from one taken in 2010, even if the underlying population hasn't changed at all. When you try to extrapolate from high-income survey countries to the full global population, you run into a selection problem. Most of the reliable data comes from Western Europe, North America, and a handful of other well-surveyed regions. Sub-Saharan Africa, parts of South Asia, and much of the Middle East are largely absent from peer-reviewed prevalence studies, either because the research doesn't exist or because it can't be published safely. So any global number is really just an imputation built on incomplete coverage. There's also the distinction between same-sex attraction, same-sex behavior, and homosexual identity. Kinsey's work from the mid-twentieth century already showed these three don't map onto each other cleanly, and modern research confirms it. A significant portion of people report same-sex attraction or experience without adopting a gay or bisexual label. How you count them changes the number by tens of millions at a global scale.
If you need a working figure for planning or analysis, the most defensible range to cite is 3 to 7 percent of the global adult population, with the understanding that the true number sits somewhere in that band and could shift upward as more countries collect better data. The wider you make the definition, the higher the number goes, and the less meaningful the comparison becomes across different studies.
Get the Full Details
