Sigmund Freud, Alfred Kinsey, E.O. Wilson, and others have suggested that social pressure suppresses natural tendencies for humans to express bisexuality, the apparent norm for one of our two closest genetic relatives, the bonobo. An analysis of data obtained from a new online sample of 1,150,938 people in 215 countries and territories (63.9% from the United States, United Kingdom, and Canada) who completed the English version of a validated questionnaire of sexual orientation lends support to this idea. A histogram of scores from 0 (exclusive opposite-sex inclinations) to 18 (exclusive same-sex inclinations) forms a near-normal distribution. Although this distribution was likely caused to some extent by sampling bias, it may also reflect the unusual honesty people show when taking online tests anonymously, as an increasing body of evidence demonstrates. We present a formal mathematical expression of a social pressure theory of sexual orientation, along with empirical evidence and computational explorations that support the theory. We also present an analysis of the new data set. Among other findings: sexual orientation labels corresponded to broad, skewed, overlapping distributions of scores. Self-labeled gays/lesbians and, to a greater extent, self-labeled straights, reported that the larger the mismatch between their sexual orientation label and their actual sexual inclinations, the more distress they felt regarding their sexual orientation, a finding that is predictable from cognitive dissonance theory. Educating the public about the true nature of sexual orientation might quell the often rancorous public debates on this topic, as well as give comfort to a large number of mislabeled people.