What do the polls tell us about the 2026 Scottish Parliament Election?
Published: 6 May 2026
6 May 2026: Ahead of tomorrow's election in Scotland, Mark McGeoghegan writes that the final polls say the broad outcome seems clear, that the SNP will remain the largest party in the Parliament, however, there are areas of uncertainty. Will they win a majority? Who will be the largest opposition party? Read in full.
6 May 2026: Ahead of tomorrow's election in Scotland, Mark McGeoghegan writes that the final polls say the broad outcome seems clear, that the SNP will remain the largest party in the Parliament, however, there are areas of uncertainty. Will they win a majority? Who will be the largest opposition party? Read in full.
Blog by Mark McGeoghegan
NOTE: all polling and prediction data is accurate as of 9am, Wednesday 6 May 2026.
As the 2026 Scottish Parliament Election campaign ends, the broad outcome seems clear: bar a massive, last-minute shift – or an unprecedented, industry-wide poll error – the SNP will remain by far the largest party in the Scottish Parliament and John Swinney will almost certainly remain as First Minister.
However, there are areas of uncertainty and unpredictability. Will the SNP win a majority? If not, will they and the Greens win enough seats combined to win a pro-independence majority? Who will be the largest opposition party? And could local and regional differences in support lead to the SNP winning fewer seats than the polls predict?
Voting intention for the 2026 Scottish Parliament Election
There are two main sources of data on how Scots intend to vote in tomorrow’s Scottish Parliament Election: Scotland-wide voting intention polls and the vote shares implied by the results of MRP models.
Standard voting intention polls use a range of data collection methods to recruit a broadly representative sample of Scots, usually around 1,000 respondents, and ask them how they would vote on both the constituency and regional list votes if a Scottish Parliament election was conducted today. They then weight their samples to match the demographics and, in some cases, past voting behaviour of the Scottish population. Their headline results factor in the likelihood that a given respondent will vote, and redistribute undecideds to predict the overall result of the election if it was held on the day respondents completed the poll.
MRPs provide a further source of voting intention data. Many in-depth explanations of how MRPs work have been published in the past month or so, and we are yet to see how they perform in aggregate as models to predict the outcome of a Scottish Parliament election, so I will not go into too much depth on how they work or how reliable they might be.
The following are the most recent voting intention estimates published by each research agency that has polled during the campaign:
|
Agency |
Vote |
SNP |
Labour |
Reform |
Cons |
Lib Dems |
Greens |
|
Ipsos |
Const. |
35% |
20% |
18% |
11% |
11% |
2% |
|
List |
26% |
15% |
18% |
10% |
11% |
17% |
|
|
Norstat |
Const. |
36% |
20% |
16% |
14% |
11% |
1% |
|
List |
28% |
17% |
17% |
14% |
10% |
12% |
|
|
More in Common |
Const. |
35% |
17% |
19% |
13% |
13% |
2% |
|
List |
26% |
15% |
22% |
11% |
12% |
11% |
|
|
Survation |
Const. |
38% |
18% |
20% |
12% |
10% |
2% |
|
List |
29% |
17% |
19% |
12% |
9% |
12% |
|
|
Stonehaven* |
Const. |
37% |
17% |
18% |
9% |
8% |
2%** |
|
List |
33% |
16% |
19% |
9% |
8% |
13% |
|
|
Find Out Now |
Const. |
35% |
14% |
16% |
9% |
10% |
2%** |
|
List |
27% |
12% |
17% |
10% |
11% |
20% |
|
|
JL Partners* |
Const. |
32% |
20% |
18% |
10% |
9% |
2%** |
|
List |
27% |
19% |
20% |
12% |
9% |
12% |
|
|
YouGov* |
Const. |
41% |
18% |
19% |
8% |
10% |
2% |
|
List |
32% |
15% |
19% |
8% |
10% |
13% |
*MRP poll
**Adjusted for the Greens standing in only six of 73 constituencies
And the following are the highest and lowest poll results for each party over the campaign period:
|
Party |
Vote |
Highest |
Lowest |
|
SNP |
Const. |
41% |
32% |
|
List |
33% |
26% |
|
|
Labour |
Const. |
20% |
14% |
|
List |
19% |
12% |
|
|
Reform |
Const. |
20% |
16% |
|
List |
22% |
17% |
|
|
Cons |
Const. |
14% |
8% |
|
List |
14% |
8% |
|
|
Lib Dems |
Const. |
13% |
8% |
|
List |
11% |
8% |
|
|
Greens |
Const. |
2% |
1% |
|
List |
17% |
11% |
Several parties have had a wide range of estimated vote shares for either or both votes throughout the campaign. The difference between the SNP’s highest and lowest constituency vote shares is 8 percentage points, the difference between a clear majority through the constituency vote and falling far short of that. Labour and Reform have roughly equivalent ceilings on both votes, but Reform has a lower floor, making it difficult to predict which would be the second-largest party. The variation in the Green list vote is the difference between a significant Green surge, potentially becoming the third-largest party, and barely making gains.
It is important to look across the polls, for this exact reason. We can do so by fitting weighted LOESS regressions to the vote share estimates of each party and using these models to predict the level of support they have in the overall voting population. LOESS regressions are a type of statistical modelling that fit models ‘locally’, so that – in this case – data from more recent polls are much more influential on the predicted vote shares than polls from longer ago. Such models would suggest that the parties have the following levels of support on the day before the 2026 Scottish Parliament Election:
|
Party |
Vote |
Predicted Vote Share |
|
SNP |
Const. |
35.8% |
|
List |
28.9% |
|
|
Labour |
Const. |
18.7% |
|
List |
16.1% |
|
|
Reform |
Const. |
17.2% |
|
List |
18.1% |
|
|
Cons |
Const. |
11.6% |
|
List |
11.4% |
|
|
Lib Dems |
Const. |
10.7% |
|
List |
10% |
|
|
Greens |
Const. |
1.7% |
|
List |
13.1% |
Using Scottish polling to predict the election outcome
Predicting seat results based on polling data is a thorny exercise, and many different approaches exist. For most of the post-war period, elections in the UK were projected using uniform national swing (UNS), in which a party losing or gaining a certain share of the vote nationally loses or gains the same share in each constituency. For example, if the Conservatives lose three points of vote share, they will lose around three points of vote share in each constituency.
Other approaches exist. Proportional swing (PS) assumes that lose and gain vote share in constituencies in proportion to their losses and gains nationally – in other words, if the Conservatives lose 10 points nationally, and this is half their vote, they will lose 30 points in a constituency in which they won 60% last time, and 10 points in a constituency where they won 20%.
More complex methods have been developed to deal with parties that have small national vote shares, but are heavily concentrated in a few areas, like the Liberal Democrats – strong transition models are a case in point. Regression models are increasingly common, using a much wider range of variables in tandem with polled voting intention to predict constituency outcomes. MRPs are an excellent example of new approaches to predicting constituency outcomes.
Again, there are many explainers available for how all these methods work1, so I won’t go into any more detail. Poll watchers will have noticed that the more complex methods remain error-prone, and almost every MRP of the campaign has produced evidently odd predictions in at least some constituencies. Much like all political polling, error is part of the game and outliers happen.
As such, we ought to look across the range of models available to us. The following are the projected seat outcomes from each research agency’s most-recent MRP during the campaign, plus a selection of models based on aggregated polling data:
|
Model |
SNP |
Labour |
Reform |
Cons |
Lib Dems |
Greens |
|
YouGov* |
62 |
17 |
19 |
7 |
8 |
16 |
|
More in Common* |
60 |
13 |
22 |
12 |
12 |
10 |
|
Stonehaven* |
66 |
13 |
21 |
7 |
8 |
14 |
|
JL Partners* |
67 |
19 |
17 |
10 |
9 |
7 |
|
DevolvedElections** |
62 |
17 |
18 |
11 |
11 |
10 |
|
Ballot Box Scotland** |
61 |
17 |
20 |
10 |
10 |
11 |
|
Mark McGeoghegan** |
59 |
18 |
19 |
10 |
12 |
11 |
|
University of St Andrews*** |
62 |
18 |
22 |
10 |
9 |
8 |
|
Median |
62 |
17 |
19.5 |
10 |
9.5 |
10.5 |
*MRP
**Proportional Swing
***Uniform National Swing
All models project that the SNP will be the largest party, with two projecting an SNP majority. The lowest seat share they are projected to win is 59 seats. Seven of the eight models project that Reform will become the second-largest party in the Scottish Parliament, with Labour projected to be the second-largest party in just one model, and to come fourth in two models (behind Reform and the Greens).
Three models project that the Conservatives will come last, with one additional model projecting that they will come joint-last with the Liberal Democrats.
What do the polls tell us about the election in Scotland?
The x-factor in the 2026 Scottish Parliament elections are the 15-20 marginal constituencies which are too-close-to-call on an individual basis. Central Belt marginals are generally contests in which the performance of Reform will determine whether the SNP or Labour will win, and marginals in the North East and South Scotland are generally contests in which the performance of Reform will determine whether the SNP or the Conservatives will win.
Certain constituencies are contests between the SNP and a party outwith the Conservatives and Labour. Banff and Buchanshire Coast is likely to be an SNP-Reform marginal, while marginals in the Highlands and Islands will be contests between the SNP and the Liberal Democrats (and, in the case of Inverness and Nairn, ex-SNP independent candidate Fergus Ewing), as will Edinburgh Northern. In Glasgow Kelvin and Maryhill, Glasgow Southside, and Edinburgh Central, the Greens are in with a shout against the SNP and Labour.
The outcomes of these marginals – which will be influenced by local and regional politics, local campaigns, and factors like turnout that can be affected by a range of unpredictable variables, like the weather on election day – will largely determine the overall outcome of the election. The SNP can win these seats on low vote shares, but with the exception of a couple of regions cannot compensate for losing them with list seats as they lack enough support on the regional list vote.
So, while the polls point to a clear SNP victory in the 2026 Scottish Parliament Election, an upset is still possible.
Author
Mark McGeoghegan is a Research Associate at the Centre for Public Policy. He is an Associate Advisory Director at Ipsos, postgraduate researcher in politics and international relations at the University of Glasgow, and a columnist in various publications.
References
[1] For UNS: https://yougov.com/en-gb/articles/11081-uniform-swing-rip
For Strong Transition Models: https://www.electoralcalculus.co.uk/blogs/strongmodel.html
First published: 6 May 2026