The state was decisively Muslim — 77.1% — yet ruled by the Hindu Dogra dynasty of Maharaja Hari Singh. Hindus were 20.1%, Sikhs 1.6%, Buddhists 1.0%.
That single statewide figure hides a fractured geography. The Kashmir Valley was over nine-tenths Muslim; the Jammu plains around the capital were Hindu-majority; Ladakh's Leh was Buddhist; the northern frontier was almost wholly Shia and Ismaili Muslim. The map below shades each district by its 1941 Muslim share.
The state ran in three divisions, and their religious arithmetic could hardly have been more different.
Kashmir Province — the Valley — was 93.5% Muslim, with a small but powerful Kashmiri Pandit minority. The Frontier Districts of Ladakh, Gilgit and Baltistan were 86.7% Muslim, leavened by Buddhist Leh. Jammu Province was Muslim-majority too, at 61.4% — but only because its mountainous west (Poonch, Mirpur) outweighed the Hindu-majority plains around Jammu city.
The Kashmir Valley held 1.73 million people, 93–95% of them Sunni Kashmiri Muslims who farmed the rice terraces and wove the shawls.
Above them sat the Kashmiri Pandits — Saraswat Brahmins, barely 4–5% of the Valley, yet the men of the revenue and judicial bureaucracy. The Pandit elite is reckoned to have owned more than 30% of the Valley's arable land. The 1933 Tenancy Act had begun handing occupancy tenants proprietary rights, but the gulf between the cultivator and the proprietor remained the defining fact of Valley life.
Jammu Province was the state in miniature — Hindu in its southeastern plains, overwhelmingly Muslim in its western hills.
Kathua was barely a quarter Muslim; Jammu district about 40%; Udhampur 43%. But Reasi was 68% Muslim, Poonch Jagir near 90%, and Mirpur — 548,000 people — roughly 80%. The map shades these districts by Muslim share: a clean east-to-west gradient that the events of 1947 would turn into a border.
Eighty to eighty-five percent of the state lived off the land, and most of them owned none of it.
Beneath the Maharaja, who was the ultimate owner under the Dogra settlement, stood the intermediaries: 396 jagirdars holding revenue grants, 2,347 mukararees drawing cash pensions, and roughly 9,000 chakdars — proprietors of estates averaging some 73 acres of irrigated paddy. Below them, Wolf Ladejinsky counted around 300,000 landless tenant families and another 250,000 part-tenants. Because the landlords were disproportionately Hindu Dogra in Jammu and Kashmiri Pandit in the Valley, while the tillers were almost all Muslim, class and creed mapped onto each other.
Roughly 9,000 landlords lost surplus land. About 450,000 acres were expropriated; some 230,000 acres passed into the ownership of cultivating tillers, the rest vesting in the state.
By April 1953, 192,652 acres had reached 160,939 tillers — an average of 1.23 acres each. One measure of the scale: of all the land redistributed across India up to 1970, the agronomist P.L. Verma reckoned that nearly half was distributed in Jammu and Kashmir alone. Ladejinsky judged the enforcement "unmistakably rigorous" — the most radical agrarian reform in non-communist South Asia.
The reform never ran across the whole state, because by 1948 the state no longer existed in one piece.
The 1947 war and the ceasefire line carved off the Muslim-majority west and north. Mirpur, the Poonch hills and Muzaffarabad became Azad Kashmir; Gilgit and Baltistan became the Northern Areas. Of the four million counted in 1941, about one million ended up on the Pakistani side. The map below colours each region by where it landed: Indian-administered, or Pakistani-administered.
In the territory that became Azad Kashmir, non-Muslims had been 12.7% of the population in 1941. By the 1951 Census of Azad Kashmir, they were 0.09%.
Of roughly 114,000 non-Muslims who had lived in those districts, Snedden found that "only a paltry 790" remained. The communal map of the former princely state was not merely redrawn by the ceasefire line — on the western side it was very nearly erased.
There is no 1951 census for Jammu and Kashmir. Every "1951" figure used by the Indian government is an arithmetic mean of the 1941 and 1961 counts.
On that basis the Indian-administered residue held about 3.25 million people in 1951; the 1961 census, the first real post-reform count, recorded 3,560,976. The state had lost its Muslim-majority frontier to Pakistan, while West Punjab refugees swelled Hindu Jammu — so the Muslim share of Indian J&K slid from roughly 77% before Partition to about 68% by 1961, a shift driven far more by lost territory than by mortality.
Land reform and demography were the same story told twice. Dispossessing the landlords meant dispossessing a largely Hindu and Pandit class; enfranchising the tillers meant enfranchising Muslims.
The reform consolidated National Conference power in the Valley, helped trigger an early wave of Pandit out-migration, and ignited the Hindu-nationalist Praja Parishad agitation in Jammu. Yet no head-count of the landlord class by religion was ever published — and a significant number of jagirdars and chakdars were themselves Muslim. The reform's communal charge was real, but the clean binary of "Hindu landlord, Muslim tenant" was always an approximation of a messier truth.
The baseline is the Census of India 1941, Vol. XXII, Jammu & Kashmir (R.G. Wreford): total 4,021,616; Muslims 77.11%, Hindus 20.12%, Sikhs 1.64%, Buddhists 1.01%. There was no 1951 census in J&K; the ~3.25 million figure is an interpolation. Azad Kashmir figures are from Iftikhar Ahmad's Census of Azad Kashmir 1951.
Land-reform numbers trace to Wolf Ladejinsky's Observations in Kashmir, Daniel Thorner's 1953 Economic Weekly essay, and the J&K Department of Information. 1947 death-toll estimates are unreconciled and stated here as a range (Snedden; Copland; Alexander–Symonds). The landlord class was never enumerated by religion. Boundaries are illustrative: Indian-administered districts are simplified from GADM v1; Azad Kashmir and Gilgit-Baltistan from geoBoundaries (OSM-derived); the shared edge approximates the Cease-Fire Line, and district Muslim shares are 1941-era approximations, not survey-grade.
A Vizmaya data story built from a research brief on the demography of Jammu and Kashmir on the eve of land reform.