How the UDF won Kerala in 2026: what the polls said and what happened

Kerala voted in a single phase on 9 April 2026, and the votes were counted on 4 May. The United Democratic Front (UDF), led by the Congress, won 102 seats. The Left Democratic Front (LDF), led by the CPI(M), won 35, and the BJP-led NDA won 3. Those alliance totals are from Wikipedia. The LDF had governed the state for ten years. V. D. Satheesan was sworn in as Chief Minister on 18 May 2026.

Three things stand out. First, the polls: the exit polls we could verify all pointed to a UDF win, but the highest UDF number in any of them was 94, and the UDF got 102. Second, the swing was mostly a Congress and IUML story. Counting parties one by one, the Congress went from 21 seats to 63, and the IUML from 15 to 22, while the CPI(M) lost 36 of its 62. Third, the turnout: about 79%, against about 74% in 2021.

Surveys published weeks before polling day are opinion polls, and they measure the mood at the time. Several of them were released in late March and early April. The table lists the ones we could read. We have not audited how any of them were done.

The polls did not agree. Most had the UDF ahead, but Mathrubhumi-Core had the LDF ahead, and The Federal described Rajneet Pulse’s 94 for the UDF as an outlier. The Federal’s own summary of Manorama-CVoter, Lok Poll, News18 and Janmat gave the UDF 67 to 86 seats and the LDF 51 to 70, and it called the contest exceptionally tight. Manorama-CVoter also asked people who they wanted as Chief Minister, as reported by Outlook India: 25% said Satheesan, 21% Pinarayi Vijayan and 17% Shashi Tharoor.

There was earlier evidence to work with. In the 2024 Lok Sabha election, our database shows the Congress winning 14 of Kerala’s 20 seats with 35.1% of the vote, and The Federal notes that the UDF led in 111 of the 140 assembly segments. A general election is not an assembly election, so this is context and not a forecast. And in the local body elections of December 2025, Wikipedia gives vote shares of 38.81% for the UDF, 33.45% for the LDF and 14.71% for the NDA, and Newsonair reports the NDA winning 50 of the 101 seats in the Thiruvananthapuram Corporation.

Exit polls ask people who have just voted, so they should be closer to the truth than opinion polls. They were published on 29 April, after voting on 9 April. The table lists the agencies whose numbers appeared in more than one source we read.

Wikipedia’s list also carries Chanakya Strategies (UDF 72 to 80), Journo Mirror (65 to 80), Vote Vibe (70 to 80), Today’s Chanakya (60 to 78) and People Insight (66 to 76). None of those put the UDF above 80 either.

On seats, the direction was right and the size was off. On vote share, two of the exit polls that gave one, Manorama-CVoter and Axis My India, had the UDF at 43% and 44%, the LDF at 38% and 39%, and the NDA at 14%. Wikipedia’s alliance totals for the result are 46.55% for the UDF, 37.64% for the LDF and 14.13% for the NDA. So on those two polls the LDF and NDA figures were close and the UDF was the one that was low. Alliance shares are counted differently by different sources, so we would not read too much into a gap of two or three points.

Sources disagree on some rows. Wikipedia gives Axis My India as 83 to 93 seats for the UDF and 47 to 57 for the LDF; Onmanorama, Deccan Chronicle and DNA India give 78 to 90 and 49 to 62, which we use. One Deccan Chronicle table had the Matrize and People’s Pulse rows labelled differently from Wikipedia and Onmanorama, so we follow the latter two. We have not audited how any agency did its survey.

Vote shares here are each party’s own, not its alliance’s. Our database does not yet hold alliance-level totals for 2026, so the alliance figures elsewhere in this piece come from Wikipedia and are not directly comparable with these rows. Parties that fought as allies in one election are not always allies in the next.

Look back a bit further and this is a big jump for the Congress. Our database has it at 22 seats in 2016 and 21 in 2021. It won 38 in 2011, 62 in 2001 and 55 in 1991. So 63 is above both of those, though the party’s vote share of 28.8% is lower than the 32.1% it had in 1991 and the 31.4% in 2001. For the CPI(M), 26 seats is its lowest since 2001 (23), and well below its 58 in 2016 and 62 in 2021. Sources also disagree on how to rank the UDF’s win: Onmanorama calls it the highest margin since 1977, while ETV Bharat and The Federal say since 1980. We have not settled that, so we do not make the claim.

Compare raw vote counts between the two elections. In 2021 the Congress polled about 52.3 lakh votes. In 2026 it polled about 62.2 lakh, roughly 9.8 lakh more. The IUML went from about 17.2 lakh to about 23.8 lakh, roughly 6.5 lakh more. The CPI(M) went from about 52.9 lakh to about 47.0 lakh, roughly 5.9 lakh fewer. The CPI lost about 1.4 lakh votes. The BJP gained about 1.1 lakh, from 23.5 lakh to 24.7 lakh.

This is arithmetic, not proof that anyone switched. Total votes polled rose by about 7.6 lakh between the two elections, from 208.3 lakh to 216.0 lakh, so every party’s count is affected by turnout, new voters and removed names as well as by people changing their choice. Without a survey of individual voters, no one can say who moved where.

The BJP’s vote share barely moved, from 11.3% to 11.4%, but its seat count went from 0 to 3. In the 2024 Lok Sabha election our database has it at 16.7% and one seat, so its assembly vote in 2026 is well below its general-election vote. Biju, quoted below, also describes the BJP’s share as largely unchanged.

One more sign of the mood: the number of voters who pressed NOTA rose from about 97,700 in 2021 to about 123,100 in 2026. We do not know why.

Because the same 140 seats were contested in 2021 and 2026, we can compare who held each one. This is by the party that won, not by alliance. The CPI(M) lost 36 of the 62 seats it won in 2021, and in every one of those 36 seats it finished second.

The CPI lost 9 of its 17 seats: 8 went to the Congress and 1 to the BJP, and it kept 8. The Kerala Congress (M) went from 5 seats to none, with 4 going to the Congress and 1 to the KEC. The Congress and the IUML held every seat they won in 2021, 21 and 15. That leaves the Congress’s 63 as 21 held plus 42 gained: 24 from the CPI(M), 8 from the CPI, 4 from the Kerala Congress (M), 2 from independents and one each from four other parties.

The three CPI(M) seats that went to independents are worth a look. According to Newslaundry and The Federal, all three winners were CPI(M) rebels: G. Sudhakaran won Ambalappuzha with 75,184 votes, by 27,935 over the CPI(M)’s H. Salam; V. Kunhikrishnan won Payyannur by 7,487 over the CPI(M)’s T. I. Madhusoodanan; and T. K. Govindan Master won Taliparamba by 12,551, with the CPI(M) second.

The BJP’s three wins were all narrow. Rajeev Chandrasekhar won Nemom by 4,978 votes over V. Sivankutty of the CPI(M), who had been Education Minister. V. Muraleedharan won Kazhakoottam by 428 over the CPI(M)’s Kadakampally Surendran. B. B. Gopakumar won Chathannoor by 4,398, with the CPI second. Nemom and Kazhakoottam were CPI(M) seats in 2021 and Chathannoor was a CPI seat. Onmanorama’s report gives the Nemom margin as 4,165; we use the Commission figure in our database. Manorama’s regional CVoter survey had projected Chandrasekhar to win Nemom.

Among the big names, V. D. Satheesan won Paravur with 78,658 votes, by 20,600 over the CPI’s E. T. Taison Master. Ramesh Chennithala won Haripad by 23,377. In Aranmula, the Congress’s Abin Varkey Kodiyattu beat the CPI(M)’s Veena George, 70,083 to 51,098. Outgoing Chief Minister Pinarayi Vijayan won Dharmadam, with 85,614 votes to the Congress candidate’s 66,367 in our database. One outlet, South First, reported a much smaller margin for him and Onmanorama said he trailed for several rounds, so we do not print a margin here. Newslaundry and Onmanorama both say thirteen ministers lost. Onmanorama names V. N. Vasavan, R. Bindu, Veena George and J. Chinchu Rani among them, and K. K. Shailaja and V. K. Prasanth among the MLAs who lost.

Overall this was not a close election. The median winning margin was 15,316 votes. Three seats were decided by fewer than 1,000 votes and 22 by fewer than 5,000, against 26 under 5,000 in 2021. Some 52 seats went by more than 20,000, against 46 in 2021, and 9 by more than 50,000. The Congress’s median margin was 16,597; it won 43 seats by more than 10,000 and 25 by more than 20,000. The IUML’s median was 30,325 and it won 16 seats by more than 20,000. The CPI(M)’s median was 10,292, and 10 of its 26 seats were under 5,000.

The IUML contested 26 seats and won 22. The biggest margin of the election was its P. K. Kunhalikutty’s 85,327 votes in Malappuram, and it holds five of the six largest wins. The other was the Congress’s Aryadan Shoukath in Nilambur, by 54,851. In Pala, an independent, Mani C Kappen, beat Jose K Mani of the Kerala Congress (M) by 2,991.

No official body explains why people vote the way they do, so the reasons below are the views of named writers. We have included only what we could read in their published pieces, and we give each one’s name or outlet.

• Fatigue with the government. The Federal’s Associate Editor Rajeev Ramachandran, writing on 4 May, cited anti-incumbency and governance fatigue after a second consecutive term, with leadership he described as centralised. Newslaundry’s Pooja Prasanna, on 11 May, also pointed to anti-incumbency, and to a view that decisions were made at the top and power was concentrated in Pinarayi Vijayan. Prof. M. R. Biju of the Central University of Kerala, in Mainstream Weekly, referred to the LDF’s autocratic style and internal revolts.
• Money and jobs. ETV Bharat’s team reported concerns about debt, inflation and livelihoods that persisted despite infrastructure projects such as Vizhinjam port and the GAIL pipeline. It quoted media observer George Abraham’s earlier warning about administrative arrogance.
• Minority consolidation. Ramachandran called the consolidation of Muslim and Christian voters the more decisive factor. He linked part of it to the amendment of the FCRA, which he described as a warning to Christians. DNA India, reporting on the Axis My India exit poll, said that poll pointed to stronger-than-usual consolidation behind the UDF and limited fragmentation of minority votes. This is what the analysts say; without a post-poll survey we cannot check it.
• The IUML. Biju described the IUML as a principal driver of the UDF’s campaign, and ETV Bharat noted its near-perfect strike rate. Our count agrees on the number: 22 wins from 26 seats contested.
• Rebels on the Left. The Federal and Newslaundry both name the CPI(M) rebels who won as independents: T. K. Govindan Master, V. Kunhikrishnan and G. Sudhakaran. Ramachandran said the rebels hurt the Left in Taliparamba and Payyannur, and he also pointed to weakness in Kannur.
• Backward-class support and the BJP. Ramachandran wrote of visible erosion of backward-class support for the Left, some of which he said benefited the BJP. Biju wrote that the BJP’s vote share was largely unchanged and that it lacked a credible alignment. Our figures agree on the share: 11.3% in 2021 and 11.4% in 2026.
• Policy and controversy. Newslaundry reported criticism of the government for what critics called soft Hindutva, citing the Ayyappa Sangamam and Sabarimala narratives. It also noted the gap in the Wayanad rehabilitation, with 178 of 410 planned houses done, and a perception that the leadership’s tone was arrogant. Outlook India wrote that late allegations, including one about Wayanad relief funds against the Congress and one about cash distribution by BJP workers in Palakkad, spread faster than they could be verified.
• The UDF’s own preparation. Biju credited early preparation by what he called TeamUDF, and ETV Bharat described Satheesan as the central architect of the campaign.

These fit the seat map. The rebel wins and the CPI(M)’s many second places are visible in the results, and the IUML strike rate is in our data. But fitting is not proving. We did not find a published post-poll survey with breakdowns by community, age or gender for this election, from Lokniti-CSDS or anyone else, and that is what would answer questions like “which communities moved?”.

Kerala also went through the Election Commission’s special revision of the voter list before the vote. According to Onmanorama, the final roll published on 21 February 2026 had 2,69,53,644 names, against 2,78,50,855 in the October 2025 base roll, so 8,97,211 fewer.

Turnout was about 79%. Our database shows 79.40%, Wikipedia says 79.53%, and DNA India reported 78.27%, so we say “about 79%”. In 2021 our database has 74.06%. The electorate we sum across seats fell from about 2.81 crore in 2021 to about 2.72 crore, while the number of votes polled rose by about 7.6 lakh. Sources also differ on the size of the electorate itself: Wikipedia gives about 2.71 crore registered voters, and the Onmanorama figure above is about 2.70 crore. We do not know why they differ.

We cannot say whether the roll revision helped or hurt any party. We found no analysis that links removed names to the result in Kerala, and we do not have names-removed data by seat ourselves.

The choice of Chief Minister took ten days from the count. Wikipedia’s page on the 2026 Kerala Chief Minister crisis names Satheesan, K. C. Venugopal and Ramesh Chennithala as the contenders. On 14 May the Congress Legislature Party elected Satheesan as its leader and the AICC announced him as Chief Minister, and on 18 May he took the oath as the 13th Chief Minister of Kerala, succeeding Pinarayi Vijayan. He had been Leader of the Opposition since 22 May 2021.

The Week reports that 20 ministers were sworn in with him, and that Mallikarjun Kharge, Rahul Gandhi, Priyanka Gandhi and the Chief Ministers of Karnataka, Telangana and Himachal Pradesh attended. Wikipedia’s page on the ministry says Ramesh Chennithala took the Home, Vigilance and Coir portfolios. For how the choice is made in general, see our explainer on how a Chief Minister is chosen, and for the list of past holders see Chief Ministers of Kerala.

• Who voted for whom, by community, age, gender or region. We found no post-poll survey, from Lokniti-CSDS or anyone else, that covers this election. The Lokniti survey that turned up in our search was from 2021.
• Why the exit polls fell short on seats. Some of the reasoning we saw is guesswork, and we did not find a post-mortem from the agencies.
• How much the roll revision, the local body results, the rebels or any single campaign controversy mattered, if at all.
• The alliance-level vote shares. We use Wikipedia’s 46.55%, 37.64% and 14.13%, and Mainstream Weekly gives 37.34% for the LDF, so they differ slightly. We are not sure which is right.
• The exact turnout and the exact size of the electorate, for the reasons given above.
• The Dharmadam margin, and how the Nemom margin was reported as 4,165 in one place and 4,978 in the Commission data.
• Whether this was the UDF’s best result since 1977 or since 1980: sources differ.
• Some exit polls (Times Now/ABP-CVoter, Jan Ki Baat) appeared only in fragments we could not verify, so we left them out.

Seats, votes, margins, seat flips and party-level vote shares are our own analysis of the results published by the Election Commission of India, from the same data that powers the Kerala 2026 election page on this site, compared with the 2021 election. Vote shares are each party’s votes divided by all votes polled, including NOTA. Where we say a party contested, that is our database’s count of its candidates; Wikipedia gives a different figure for the CPI(M) (77, against our 81). Alliance totals, opinion polls, exit polls and campaign details come from the outlets named, and analyst views are credited to the person or outlet that wrote them. If you spot a number that does not match the Commission’s, tell us on the contact page.