DMK won 133 of 234 Assembly seats in 2021. In 2026 they won 59. Tamil Nadu replaced its ruling party, and the team most analysts had been calling well-organised, well-funded, and incumbency-protected lost three out of every four seats it had defended.
The interesting question is not whether DMK fell. They obviously did. The question is where. Because the collapse was not uniform. It was concentrated in exactly the regions where DMK had been the strongest one cycle earlier.
The headline-friendly version of this story is M.K. Stalin losing his own seat in Kolathur, which has already been covered in detail. The structural version is the one below: a region-by-region accounting of where the seat math came apart, and what 24.3% statewide vote share buys you under first-past-the-post.
The collapse hits hardest where DMK was strongest
In 2021, DMK polled above 40% in three regions: Central (46.2%), Chennai & Suburbs (42.9%), and the North (42.2%). These were the fortresses.
In 2026, those same three regions dropped 17.4, 19.4, and 18.6 percentage points of vote share respectively. The places DMK had been strongest delivered the steepest falls.
The pattern is consistent across the dataset. The correlation between a region’s 2021 DMK vote share and the size of its 2026 share drop is +0.67. The taller they stood, the further they fell. Regions where DMK was already weak (Deep South at 29.2%, Kongu at 34.6%) shed only 7 to 10 percentage points. Their floor was closer to the ground.
This matters because it overturns the easier narrative that DMK lost because of marginal-seat erosion. They did not. They lost because their core broke.
The Chennai region gave DMK 31 of its 37 seats in 2021. In 2026 it gave them 2. The fortress was the wave’s first stop.
Region by region
| Region | ACs | DMK won 2021 | DMK won 2026 | Vote share 2021 | Vote share 2026 | Change | Of seats lost, to TVK | To AIADMK | To other |
|---|---|---|---|---|---|---|---|---|---|
| Chennai & Suburbs | 37 | 31 | 2 | 42.86% | 23.42% | -19.44 pp | 28 | 1 | 0 |
| North | 21 | 15 | 3 | 42.21% | 23.61% | -18.60 pp | 5 | 7 | 0 |
| Central | 23 | 20 | 8 | 46.17% | 28.75% | -17.42 pp | 8 | 3 | 1 |
| Krishnagiri Belt | 11 | 2 | 1 | 31.54% | 16.15% | -15.39 pp | 0 | 2 | 0 |
| Northeast Coast | 20 | 12 | 7 | 37.11% | 22.19% | -14.92 pp | 1 | 4 | 2 |
| Madurai Region | 21 | 12 | 7 | 39.74% | 29.08% | -10.66 pp | 7 | 1 | 0 |
| Kongu (West) | 46 | 11 | 7 | 34.63% | 24.30% | -10.33 pp | 7 | 3 | 0 |
| Cauvery Delta | 18 | 11 | 9 | 37.62% | 28.65% | -8.97 pp | 2 | 0 | 2 |
| Deep South | 37 | 19 | 15 | 29.20% | 22.00% | -7.20 pp | 7 | 1 | 1 |
| State | 234 | 133 | 59 | 37.70% | 24.31% | -13.39 pp | 65 | 22 | 6 |
Two extremes anchor the table.
Chennai & Suburbs is the worst single result. DMK won 31 of 37 seats here in 2021. They won 2 in 2026. Of the 29 seats they lost, 28 went directly to TVK. The party’s vote share fell from 42.86% to 23.42%. This is the M.K. Stalin region, the cabinet’s home base, and a stretch DMK has organised on every ward for two decades. It produced two MLAs.
Krishnagiri Belt is the most telling. DMK was not strong there in 2021 (31.5%) and crashed to 16.15% in 2026, the lowest regional vote share they polled anywhere in the state. They lost both their 2021 seats in this region to AIADMK. Krishnagiri Belt is the one region where the seat math fell apart not because TVK swallowed the seats but because AIADMK held on. The DMK-versus-TVK story everyone is telling does not apply equally in every part of Tamil Nadu.
Cauvery Delta is the rump. DMK lost only 2 seats (11 to 9) and a modest 9 points of vote share. With 28.65% they polled their second-best share of any region. The Delta is where the DMK organisation, the agrarian economy, and the alliance still added up to a working coalition. It is what DMK looks like when nothing else goes wrong.
Where the seats actually went
Of the 93 seats DMK held in 2021 and lost in 2026:
- 65 went to TVK. This is the headline number. Roughly 7 of every 10 DMK seats that flipped landed in TVK’s column.
- 22 went to AIADMK. Most of these are in the North, the Northeast Coast, and the Krishnagiri Belt, where AIADMK’s organisational footprint never fully eroded and the anti-DMK vote rallied behind the older opposition.
- 3 went to PMK (Gingee, Vikravandi, Jayankondam) and the NDA alliance, a small but specific Vanniyar consolidation in the Salem-Cuddalore border districts.
- 3 went to smaller parties (IUML, AMMK, CPI-M), all narrow wins in specific seats where local arithmetic broke DMK’s way.
DMK did pick up 19 new seats they had not held in 2021. Of those, 15 came from AIADMK, 2 from BJP, 1 from VCK, and 1 from Congress. These are mostly in the south, where AIADMK collapsed and DMK was the next-largest organisation standing. But the gross-versus-net math, 93 out, 19 in, makes the floor clear.
What this is, and isn’t
This is a region-level accounting of where DMK’s seat majority unraveled. It does not explain why the wave hit. The structural drivers (anti-incumbency, alliance fatigue, Vijay’s brand, cadre defections) are layered and seat-specific. The full per-AC story lives in the TN 2026 Explorer dashboard, with the constituency map, the 2021→2026 swing table for every seat, and the 9-region breakdown for all parties.
The pattern that stays consistent across geography is that incumbency was a liability and not a moat. The 12 most decisive defeats of sitting MLAs are documented in the incumbents post. The full demographic replacement is in the new assembly profile. The TVK side of the same story, with the regional asymmetry that produced 108 seats, is in the TVK post.
DMK is now the second-largest party in the Tamil Nadu Assembly. They will lead the opposition with 59 seats. But the structural picture is sharper than the seat count: their core regions are gone, and the path back runs through Chennai, Central, and the North, where they bled the most.
Explore the data
Released under CC BY 4.0. Suggested citation: Anand Raj, “How DMK Lost Tamil Nadu, Region by Region.” ndranandraj.com, May 2026.