Every tradition covered in the previous twenty-five issues once had a spell as a "new religion" — small, founder still living, membership almost entirely converts, and widely regarded by the neighbours as not a real religion at all. Studying new religions means watching at close range something normally reconstructed from documents: how a religion starts, and what it takes to outlive the first generation.
"Men may believe what they cannot prove. They may not be put to the proof of their religious doctrines or beliefs."— United States v. Ballard (1944); the case established that a court may test sincerity of belief, but not the truth of what is believed
Start with the label. In the early twentieth century Troeltsch and Weber used the church–sect pair to distinguish large institutionalised bodies from small high-tension groups — a neutral analytic tool. By the 1960s and 70s in the US and Western Europe, cult had curdled in popular usage into a synonym for "dangerous fake religion." So scholars started over: new religious movement (NRM) spread in the late 1970s precisely because it prejudges nothing. Barker's working criterion is plain — young, and first-generation membership consisting mainly of converts. The convert's intensity and sense of rupture is where nearly every other feature originates.
The restorationist wave (19th-century America). Out of the embers of the awakenings came a cluster of groups claiming to restore the original church: the Latter-day Saint movement (1830), Seventh-day Adventists (1863), the Bible Students who later became Jehovah's Witnesses (1870s), Christian Science (1879). The shared logic: existing churches have drifted, a reset is required. A parallel line was the Theosophical Society (1875), which carried karma, rebirth and spiritual evolution into Western vocabulary.
The postwar Japanese wave. After 1945, religious freedom and the collapse of the old order combined, and a set of mostly prewar-founded organisations expanded fast — the period is often called the "rush hour of the gods." Soka Gakkai is the largest of them.
The counterculture wave (1954–1970s). Scientology (1954), the International Society for Krishna Consciousness (1966), the Unification Church's westward spread, communal experiments of every kind — aimed at educated baby boomers. Conflict was sharpest here, because membership often meant leaving school and family behind.
The de-institutionalising wave (1980s onward). New Age and the human potential movement pulled spirituality out of congregations and repackaged it as workshops and courses; after 2000, platforms and algorithms took over distribution. The signature of this wave is that there is no membership roll: plenty of practice, zero affiliation.
How many groups. The INFORM database, founded by Barker in 1988, holds over 2,000 groups, on which basis Europe alone plausibly contains upward of two thousand; Melton, using a narrower definition that excludes human-potential groups, can supply information on close to 1,000 American groups. The threefold gap is not a counting error — it is where each draws the line around "religion."
How many people depends on how you ask. Jehovah's Witnesses are the clearest case, because the organisation itself publishes four figures at once: for the 2025 service year, 9.05 million average monthly publishers, a peak of 9.21 million, 304,643 baptised during the year — and worldwide Memorial attendance of 20.6 million. The outer ring is more than twice the inner one.
Self-reported versus independent estimates. The Church of Jesus Christ of Latter-day Saints reported 17.89 million members at the end of 2025 — a baptismal-roll figure that includes long-inactive members and is not reduced by attrition; self-identification in national surveys runs well below the roll. Soka Gakkai reports 8.27 million member households in Japan plus roughly 3 million abroad, while the scholar Levi McLaughlin estimated in 2018 that the real Japanese figure may be under 4 million. Scientology has long claimed millions; independent estimates are in the tens of thousands (the World Christian Encyclopedia's 2020 edition gives about 70,000). Estimates for Cao Dai in Vietnam range from 2.4 to 4.4 million, and up to roughly 6 million counting all branches. Most of the gaps are explicable: rolls that never shrink, household-to-person conversions, counting anyone who took a course.
Then the large uncounted layer. In Pew's 2023 survey, 22% of US adults described themselves as spiritual but not religious; four in ten believe in psychics, four in ten that spiritual energy resides in physical objects, 33% in reincarnation, 29% in astrology, with about six in ten accepting at least one — and these beliefs are common among the religiously affiliated and unaffiliated alike. This layer has no congregation, no canon and no initiation, so organisation-based statistics cannot reach it.
No single geographic centre. The restorationist bodies now grow mainly in sub-Saharan Africa, Latin America and the Pacific islands. The one durable regularity: the looser the regulation of religion and the greater the population movement, the more new groups appear.
New religions share no theology — the only reason to consider them together is sociological. What can be generalised are a few recurring structures.
Wallis's 1984 three-way division remains the textbook standard, and its virtue is that it looks only at a group's relation to the surrounding world, never at whether its doctrine is true.
The first box draws the most attention and holds the fewest people; the third holds the most, precisely because it is unremarkable. The movement researchers watch is from the first toward the third — Weber's routinisation of charisma, observable here in real time.
Misreading: new religions control members by brainwashing.
The brainwashing hypothesis was ruled scientifically unsupported and inadmissible in US courts, and mainstream scholarship does not use it as an explanation. Empirical work gives the opposite picture: a small share of contacts join, a large share of joiners leave, and recruitment travels along existing personal ties. None of which denies that coercion, fraud, sexual exploitation and physical harm occur — those are measured by ordinary law, independent of what anyone believes.
Misreading: "cult" is an objective category you can classify with.
Scholars do not use it analytically, because it packs three separate things into one word: I disagree with the doctrine, the group is small and unfamiliar, and the group has actually broken the law. The first two carry no evidential weight; the third is handled by ordinary criminal law. What is workable is a conduct standard, not a doctrinal one.
Misreading: new religions are dangerous as a class.
The major violent episodes are few and clustered: Jonestown (1978, 918 dead), Waco (1993, 76), the Order of the Solar Temple (1994–97, 74 across several events), the Tokyo subway sarin attack (1995, 13 dead and thousands injured), Heaven's Gate (1997, 39). Against 2,000-plus documented groups, that is a very small proportion. Researchers mostly explain these through a conflict spiral: interaction between a group and the outside world — media, relatives, law enforcement — escalates step by step until both sides expect the worst, and only then does apocalyptic doctrine become an operational programme. Doctrine alone is not a predictor.
Misreading: new religions never last.
Most do vanish within a generation, which is exactly what makes the survivors interesting. The Bahá'í Faith, Sikhism, Islam and Christianity all met the definition of a "new religion" at their own starting points. Today's major traditions are simply the new religions that lived.
This issue runs the previous twenty-five in reverse. The early Christians of Day 2 were a suspect little group in Roman eyes; the Sikhism of Day 12 was wholly new in the sixteenth century; the Bahá'í Faith of Day 18 is under two hundred years old, and gets its own issue here precisely because it cleared the first-generation hurdle. The only difference is time: everything that now counts as a "tradition" is a new religion that did not die.
Continuous with Days 23 and 25. Day 23 covered how a canon forms; new religions compress that process into a few decades, so how new revelation is set beside older scripture, who fixes the version, and when oral material is written down can all be watched directly. Day 25 covered institutional loss; this issue supplies the other half — demand did not disappear, it moved to the side with no membership roll.
Two legal routes, each with a cost. Since Ballard, US courts do not evaluate doctrine and address conduct only; France's 2001 About-Picard law and the subsequent MIVILUDES mechanism established dedicated monitoring and use the category of dérives sectaires, sectarian deviance. The first risks intervening too late, the second risks treating the unfamiliar as dangerous. The shared floor is roughly this: belief is not subject to review; conduct is subject to ordinary law.
This page presents the research frameworks, counting conventions and contemporary situation of new religious movements from a neutral, comparative standpoint. It makes no judgement on the truth of any group's doctrine and does not evaluate serving religious figures. Membership figures are self-reported by the groups or estimated by research bodies; they vary widely with definition and method, and the counting frame is noted wherever possible.