European Parliament seat projection (September 2025): Further losses for the centre, gains for the left and the far right

Deutsch

By Manuel Müller

	Left	G/EFA	S&D	RE	EPP	ECR	PfE	ESN	NI	other
EP2024	46	53	136	77	188	78	84	25	33	–
EP today	46	53	136	75	188	79	85	27	31	–
July 25 (B)	51	44	124	73	181	80	99	36	20	12
Sep. 25 (B)	55	40	123	66	176	85	105	37	21	12
Sep. 25 (D)	56	40	125	68	177	89	107	37	21	–

Baseline scenario,
as of 1 September 2025.
(Click to enlarge.)

Dynamic scenario,
as of 1 September 2025.
(Click to enlarge.)

The summer recess in Brussels and Strasbourg has come to an end. Committee and parliamentary group meetings have resumed in the European Parliament in recent days, and the first major event of the new season – the State of the European Union address by Commission President Ursula von der Leyen (CDU/EPP) – will take place already next Wednesday.

For MEPs, this is also an important opportunity to take stock of their own situation: What is the current state of the von der Leyen coalition between the European People’s Party (EPP), the Social Democrats (S&D) and the Liberals (RE)? Have the wounds from early summer, when S&D and RE issued an “ultimatum” to the Commission president, left any scars? Has the failed vote of no confidence in July, in which the political centre groups voted (almost) unanimously for von der Leyen, welded the coalition back together? Or does the fear that the increasingly right-wing EPP could abandon its pro-European partners in future votes and instead rely on the support of far-right groups still prevail?

Grand coalition at a record low

One factor that certainly dampens hopes for a successful restart of the von der Leyen coalition is its performance in Europe-wide polls. Since the European elections in June 2024, the ratings of the centrist groups (EPP, S&D, RE and Greens/EFA) have fallen almost continuously, while the far right (ECR, PfE, ESN) and the Left group have gained ground.

This development had reached a historic point already in the last seat projection before the summer recess: Since the European Parliament was established, the far-right bloc had never been stronger and the parties forming the von der Leyen coalition had never been weaker. Now, in the first projection after the summer, there are no signs of this trend reversing. Quite the opposite, in fact: it seems to have accelerated. Centrist parties are losing ground again, while the far right continues to gain. If elections were held today, the European Parliament would be, at the same time, the most polarised and the most right-wing it has ever been.

EPP: heavy losses

The losses were particularly severe for the EPP, which over the summer lost support in member states of very different sizes and geographical locations – from Estonia to Germany, from Austria to Spain. In each of these countries, its poll ratings fell by only a few percentage points. In some cases, these may just be random fluctuations, especially since fewer polls are usually conducted during the summer break than at other times of the year. Overall, however, the trend is undoubtedly worrying for the EPP, as its ratings were mediocre at best even in the months before the summer break.

Only a few countries are seeing EPP member parties gain ground. In the Netherlands, for example, the CDA has significantly closed the gap on the far-right PVV (PfE) and the red-green PvdA-GL (S&D-G/EFA) alliance ahead of the national parliamentary elections in October. In Poland, Donald Tusk’s ruling PO party is recovering, albeit at the expense of its national coalition partners. In Lithuania, the TS-LKD is now the strongest force in the country. Nevertheless, all of this does not prevent the EPP from falling to 176 seats in the current projection (–⁠5 since July), which is its worst result in this legislative term.

S&D weaker than ever

The developments were hardly any better for the centre-left S&D group (123 seats/–⁠1). It too has recently recorded losses in polls in numerous member states (including Germany, Spain and Lithuania), but has also seen slight recoveries elsewhere (e.g. in Greece, Portugal and Hungary).

Overall, the S&D’s polling in recent weeks has been a mixed bag, and it certainly has not lost quite as much ground as the EPP. Even before the summer, however, the Social Democrats were at an all-time low in the baseline scenario of the projection – and now they have once again fallen below that level.

Liberals in decline

The biggest loser in recent weeks is the centrist-liberal RE group, whose decline has been ongoing for three years and has recently intensified. This is partly due to waning support for the long-standing governing parties in the Netherlands (VVD) and in France (RE). Additionally, the disunity and fragmentation of the various small liberal parties in Italy (Italia Viva, Azione and Più Europa) means that none of them would currently reach the national 4% threshold.

The Liberals are seeing a slight recovery only in Lithuania, where the small LP party would now just manage to win one seat. Overall, the RE group now stands at 66 seats (–⁠7) in the baseline scenario of the projection. This is its worst result since the 2014 European elections and thus since regular seat projections for the European Parliament began.

Greens still struggling

In the baseline scenario of the projection, the combined total of EPP, S&D and RE is now 365 seats – four more than the mathematical requirement for an absolute majority. In the dynamic scenario, the von der Leyen coalition fares slightly better with 370 seats. Even so, given the notoriously low level of party discipline in the European Parliament, a reliable pro-European majority would now need the support of the Greens/EFA group.

However, the Greens are also struggling with poor poll ratings themselves. Having made strong gains in the July projection, their group has now fallen back to 40 seats (–⁠4) – one of its worst results in the last three years.

Still, the Greens can take some comfort from the fact that the losses in recent weeks were mainly due to the French Écologistes, who are currently projected below the national 5% threshold – a result that is highly unlikely to happen in an actual European election. In general, Green parties tend to mobilise more effectively in European elections than in national ones, so their results typically outperform their (national) poll ratings. This is particularly true in France, where the Écologistes usually run as part of a broad left-wing alliance in national elections due to the first-past-the-post system, whereas in European elections they are more visible as an independent force.

As these effects are difficult to quantify, they are not taken into account in the projection. However, if the next European election really were to take place this week, the Greens would probably perform better than the projection suggests.

Left gains ground again

Conversely, the Left group is gaining ground, with its member parties making progress particularly in the large member states of France, Italy and Spain. While these changes are mostly small and could also be attributed to random fluctuations in the polls, the gains made in recent weeks build on similar developments in previous months that have now been consolidated.

In view of the crisis facing the centrist grand coalition, a growing number of European voters seem to be placing their hopes in the Left, which presents itself as the antithesis of the rising far-right parties. In the baseline scenario of the current seat projection, the Left group achieves 55 seats (+⁠4), which is nine more than in the 2024 European election and its best result in over three years.

Far-right camp at all-time high

Finally, the three far-right groups – the ECR, the PfE and the ESN – are experiencing significant gains once again. For the first time, the three groups combined would now account for over 30 per cent of the Parliament’s seats. Each of them is performing better than ever before in the seat projections.

Similar to the losses suffered by centrist parties, the recent gains made by far-right parties are spread widely across the EU member states. The EKR faction gained slightly in Italy, Poland, Romania, Greece and Sweden, climbing to 85 seats (+⁠5). The PfE faction saw gains in France, Spain, Austria and Czechia, among others, offsetting slight losses in Poland (105 seats/+⁠6). Finally, within the small ESN group, the German AfD in particular made significant gains, offsetting losses in Poland and Czechia (37 seats/+⁠1).

As with the other political groups, some minor changes in the number of seats held by individual national parties may be due to random fluctuations in the polls. However, the trend here has also been pointing in the same direction for several months now. The ESN group has now improved its seat projection for the seventh time in a row, increasing its number of seats by around half since the European elections in the summer of 2024. Similarly, the PfE has increased its number by around a quarter during the same period.

Non-attached and “other” parties

There have been few changes among the non-attached parties in the European Parliament. Two left-wing conservative parties, Germany’s BSW and Czechia’s Stačilo, have gained slightly, while Greece’s left-wing populist PE falls back. Overall, the non-attached parties now hold 21 seats (+⁠1).

Among the “other” parties (which are not currently represented in the European Parliament and do not belong to any European party, meaning they cannot be clearly attributed to any political group), the Eurosceptic, right-wing populist Dutch party JA21 is gaining ground, while the Latvian far-right party NA is losing it. The Slovenian centre-right party Demokrati is back to the table, while the right-wing populist, anti-vaccination party Resni.ca, also from Slovenia, would not win any seats in the European Parliament now. Overall, the “others” continue to hold 12 seats (±⁠0).

The overview

The following table breaks down the projected distribution of seats by individual national parties. The table follows the baseline scenario, in which each national party is attributed to its current parliamentary group (or to the parliamentary group of its European political party) and parties without a clear attribution are labelled as “other”.

In contrast, the dynamic scenario of the seat projection assigns each “other” party to the parliamentary group to which it is politically closest. In addition, the dynamic scenario also takes into account likely future group changes of parties that are already represented in the Parliament. In the table, the changes from the baseline to the dynamic scenario are indicated by coloured text and in the mouse-over text. The mouse-over text also lists any alternative groups that the party in question might plausibly join.

In the absence of pan-European election polls, the projection is based on an aggregation of national polls and election results from all member states. The specific data basis for each country is explained in the small print below the table. For more information on European parties and political groups in the European Parliament, click here.

	Left	G/EFA	S&D	RE	EPP	ECR	PfE	ESN	NI	other
EP2024	46	53	136	77	188	78	84	25	33	–
EP today	46	53	136	75	188	79	85	27	31	–
July 25 (B)	51	44	124	73	181	80	99	36	20	12
Sep. 25 (B)	55	40	123	66	176	85	105	37	21	12
Sep. 25 (D)	56	40	125	68	177	89	107	37	21	–
	Left	G/EFA	S&D	RE	EPP	ECR	PfE	ESN	NI	other
DE	9 Linke 1 Tier	10 Grüne 3 Volt	12 SPD	3 FDP 3 FW	23 Union 1 Familie 1 ÖDP			23 AfD	4 BSW 2 Partei 1 PdF
FR	11 LFI		11 PS	14 RE	11 LR		34 RN
IT	11 M5S 2 SI	4 EV	19 PD		6 FI 1 SVP	25 FdI	8 Lega
ES	2 Sumar 3 Pod 1 Bildu	2 Sumar 1 ERC	17 PSOE	1 PNV	22 PP	1 SALF	10 Vox		1 Junts
PL			3 Lewica		19 KO	20 PiS	5 Konf	6 Konf
RO			7 PSD	4 USR 1 PMP	6 PNL	15 AUR
NL	1 SP	3 GL	4 PvdA	4 VVD 2 D66	6 CDA 1 BBB		8 PVV			2 JA21
BE	3 PTB	1 Groen	2 Vooruit 2 PS	2 MR 2 LE	2 CD&V 1 CSP	4 N-VA	3 VB
CZ		2 Piráti			2 STAN 1 TOP09 1 KDU-ČSL	3 ODS	8 ANO 1 AUTO	2 SPD	1 Stačilo
EL	1 Syriza		4 PASOK	1 KD	6 ND	3 EL	1 FL		2 PE 2 KKE	1 MéRA
HU			1 DK		10 TISZA		9 Fidesz	1 MHM
PT		1 Livre	6 PS	2 IL	7 AD		5 Chega
SE	2 V	1 MP	8 S	1 C	4 M	5 SD
AT		2 Grüne	4 SPÖ	2 Neos	4 ÖVP		8 FPÖ
BG			2 BSP	2 PP	5 GERB 1 DB			3 V	3 DPS-NN	1 MECh
DK	1 Enhl.	3 SF	4 S	2 V	2 LA 1 K	1 DD	1 DF
SK				4 PS	2 Slov 1 KDH	1 SaS		2 REP	3 Smer 2 Hlas
FI	2 Vas	1 Vihreät	4 SDP	3 Kesk	3 Kok	2 PS
IE	4 SF			4 FF	4 FG					2 SD
HR		2 Možemo	4 SDP		5 HDZ					1 Most
LT		2 DSVL	2 LSDP	1 LS 1 LP	3 TS-LKD	1 LVŽS				1 NA
LV		1 Prog			1 JV	2 NA 1 LRA	2 LPV			1 ZZS 1 ST!
SI			1 SD	3 GS	4 SDS					1 D
EE			1 SDE	1 RE 1 KE	2 Isamaa		2 EKRE
CY	1 AKEL		1 DIKO		2 DISY	1 ELAM				1 ALMA
LU		1 Gréng	1 LSAP	2 DP	2 CSV
MT			3 PL		3 PN

Timeline (baseline scenario)

	Left	G/EFA	S&D	RE	EPP	ECR	PfE	ESN	NI	other
25-09-01	55	40	123	66	176	85	105	37	21	12
25-07-01	51	44	124	73	181	80	99	36	20	12
25-05-19	49	40	130	76	179	79	100	35	23	9
25-03-24	52	41	131	73	177	79	99	33	24	11
25-01-27	48	43	130	81	185	77	93	29	24	10
24-12-02	43	41	131	83	186	73	100	27	24	12
24-10-07	44	41	136	79	186	74	96	26	29	9
24-08-12	44	45	137	77	191	73	88	25	31	9
EP 2024	46	53	136	77	188	78	84	25	33	–

Timeline (dynamic scenario)

	Left	G/EFA	S&D	RE	EPP	ECR	PfE	ESN	NI	other
25-09-01	56	40	125	68	177	89	107	37	21	–
25-07-01	52	44	126	75	181	84	101	37	20	–
25-05-19	49	40	132	78	179	82	101	36	23	–
25-03-24	52	41	132	74	179	82	103	33	24	–
25-01-27	49	43	132	82	185	80	96	29	24	–
24-12-02	43	42	133	82	186	77	104	27	26	–
24-10-07	46	41	137	79	187	77	97	26	30	–
24-08-12	45	46	138	78	191	76	89	25	32	–
EP 2024	46	53	136	77	188	78	84	25	33	–

The “EP 2024” line indicates the distribution of seats as of July 16, 2024, when the European Parliament was constituted following the election in June 2019.
Overviews of older seat projections from previous legislative terms can be found here (2014-2019) and here (2019-2024).

Attribution of national parties to parliamentary groups

Baseline scenario: The projection assigns parties that are already represented in the European Parliament to their current parliamentary group. National parties that are not currently represented in the European Parliament but belong to a European political party, are attributed to the parliamentary group of that party. In cases where the members of a national electoral list are expected to split up and join different political groups after the election, the projection uses the allocation that seems most plausible in each case (see below). Parties for which the allocation to a specific parliamentary group is unclear are classified as “other” in the baseline scenario.

According to the rules of procedure of the European Parliament, at least 23 MEPs from at least a quarter of the member states (i.e. 7 out of 27) are required to form a parliamentary group. Groupings that do not meet these conditions would therefore have to win over additional MEPs in order to be able to constitute themselves as a parliamentary group.

Dynamic scenario: In the dynamic scenario, all “other” parties are assigned to an already existing parliamentary group (or to the group of non-attached members). In addition, the dynamic scenario also takes into account other group changes that appear politically plausible, even if the respective parties have not yet been publicly announced them. To highlight these changes from the baseline scenario, parties that are assigned a different parliamentary group in the dynamic scenario are marked in the colour of that group. Moreover, the name of the group appears in the mouse-over text. Since the attributions in the dynamic scenario are partly based on a subjective assessment of the political orientation and strategy of the parties, they can be quite uncertain in detail. From an overall perspective, however, the dynamic scenario may be closer to the real distribution of seats after the next European election than the baseline scenario.

The full names of the political groups and of the national parties appear as mouse-over text when the mouse pointer is held still over the name in the table. In the case of “other” parties and parties that are likely to change group after the next European elections, the mouse-over text also lists the groups that the party might join. The group to which the party is assigned in the dynamic scenario is listed first.

Data source

If available, the most recent poll of voting intentions for the European Parliament is used to calculate the seat distribution for each country. In case that more than one poll has been published, the average of all polls from the two weeks preceding the most recent poll is calculated, taking into account only the most recent poll from each polling institute. The cut-off date for taking a survey into account is the last day of its fieldwork, if known, otherwise the day of its publication.

For countries where the last specific European election poll was published more than a fortnight ago or where significantly fewer polls for European than for national parliamentary elections were published in the last two weeks, the most recent available poll for the national parliamentary election or the average of all national or European parliamentary polls from the two weeks preceding the most recent available poll is used instead. For countries where there are no recent polls for parliamentary elections, polls for presidential elections may be used instead, with the presidential candidates’ polling figures assigned to their respective parties (this concerns France and Cyprus in particular). For member states for which no recent polls can be found at all, the results of the last national or European elections are used.

As a rule, the national poll results of the parties are directly projected to the total number of seats in the country. For countries where the election is held in regional constituencies without interregional proportional compensation (currently Belgium and Ireland), regional polling data is used where available. Where this is not the case, the number of seats is calculated for each constituency using the overall national polling data. National electoral thresholds are taken into account in the projection where they exist.

In Belgium, constituencies in the European election correspond to language communities, while polls are usually conducted at the regional level. The projection uses polling data from Wallonia for the French-speaking community and polling data from Flanders for the Dutch-speaking community. For the German-speaking community, it uses the result of the last European election (1 seat for CSP).

In countries where it is common for several parties to run as an electoral alliance on a common list, the projection makes a plausibility assumption about the composition of these lists. In the table, such multi-party lists are usually grouped under the name of the electoral alliance or of its best-known member party. Sometimes, however, the parties of an electoral alliance split up after the election and join different political groups in the European Parliament. In this case, the parties are listed individually and a plausibility assumption is made about the distribution of list places (usually based on the 2024 European election results). This includes the following cases: Spain: Sumar: Sumar (place 1 and 6 on the list), CatComù (2), Compromís (3), IU (4) and Más País (5); Ahora Repúblicas: ERC (1, 4), Bildu (2) and BNG (3); CEUS: PNV (1) and CC (2); Romania: ADU: USR (1-2, 4-5, 7-9), PMP (3) and FD (6); Netherlands: PvdA (1, 3, 5 etc.) and GL (2, 4, 6 etc.); Czechia: Spolu: TOP09 (1, 3, 5 etc.) and KDU-ČSL (2, 4, 6 etc.); Stačilo: Stačilo (1, 3, 5 etc.) and SOCDEM (2, 4, 6 etc.); Hungary: DK (1-4, 6, 8), MSZP (5) and PM (7). When the election comes closer and the parties announce their candidates, the projection uses the distribution on the actual list instead. In some countries, the exact distribution of seats within an electoral alliance depends on preference votes and/or regional constituency results, so that only a plausible assumption can be made in advance. This concerns the following cases: Italy: AVS: SI (1, 3) and EV (2, 4); Poland: Konfederacja: NN (1, 3, 5 etc.), RN (2, 4, 6 etc.). In France, some polls combine PS (S&D), EELV (G/EFA), PCF (Left) and sometimes LFI (Left); in these cases, the polling figures are divided between the parties according to the ratio in the last poll in which they were reported separately. In Czechia, some polls combine ODS (ECR), TOP09 and KDU-ČSL (both EPP); in these cases, two thirds of the seats are allocated to the ODS and one third to the alliance of TOP09 and KDU-ČSL. In Italy, a special rule allows minority parties to enter the Parliament with only a low number of votes, provided they form an alliance with a larger party. The projection assumes such an alliance between FI and the SVP.

Since there is no electoral threshold for European elections in Germany, parties can win a seat in the European Parliament with less than 1 per cent of the vote. Since German polling institutes do not usually report values for very small parties, the projection includes them based on their results in the last European election (3 seats each for Volt and FW, 2 seats for Partei, 1 seat each for Tierschutzpartei, ÖDP, Familienpartei, and PdF). If a small party achieves a better value in current polls than in the last European election, the poll rating is used instead. For Slovenia, the seat projection assumes that MEP Vladimir Prebilič will run for the Vesna party as in 2024; if polls show values for a hypothetical party of Prebilič’s own, these are attributed to Vesna.

The following overview lists the data source for each member state. The dates refer to the last day of the fieldwork; if this is not known, to the day of publication of the polls:

Germany: national polls, 18-29/8/2025, source: Wikipedia.
France: national polls, 29/8/2025 (ratio between centre-left parties: 4/6/2025), source: Wikipedia.
Italy: national polls, 25/8/2025, source: Wikipedia.
Spain: national polls, 28-29/8/2025, source: Wikipedia.
Poland: national polls, 13-25/8/2025, source: Wikipedia.
Romania: national polls, 10-23/7/2025, source: Wikipedia.
Netherlands: national polls, 25-26/8/2025, source: Wikipedia.
Belgium, Dutch community: polls for the national parliamentary election in Flanders, 23/5-3/6/2025, source: Wikipedia.
Belgium, French community: polls for the national parliamentary election in Wallonia, 23/5-3/6/2025, source: Wikipedia.
Belgium, German community: Results of the European Parliament election, 9/6/2024.
Czechia: national polls, 22-26/8/2025, source: Wikipedia.
Greece: national polls, 25/8/2025, source: Wikipedia.
Hungary: national polls, 6-19/8/2025, source: Wikipedia.
Portugal: national polls, 18/8/2025, source: Wikipedia.
Sweden: national polls, 17/8/2025, source: Wikipedia.
Austria: national polls, 17-30/7/2025, source: Wikipedia.
Bulgaria: national polls, 14-23/7/2025, source: Wikipedia.
Denmark: national polls, 20-24/8/2025, source: Wikipedia.
Slovakia: national polls, 10-19/8/2025, source: Wikipedia.
Finland: national polls, 5-17/8/2025, source: Wikipedia.
Ireland: national polls, 1/8/2025, source: Wikipedia.
Croatia: national polls, 25/8/2025, source: Wikipedia.
Lithuania: national polls, 25/7/2025, source: Wikipedia.
Latvia: national polls, July 2025, source: Wikipedia.
Slovenia: national polls, 13-25/8/2025, source: Wikipedia.
Estonia: national polls, 11-19/8/2025, source: Wikipedia.
Cyprus: national polls, 28/6-8/7/2025, source: Wikipedia.
Luxembourg: national polls, 24/4/2024, source: Wikipedia.
Malta: national polls, 6/6/2025, source: Wikipedia.

Pictures: all graphs: Manuel Müller.