Boundary Agreements are Here to Stay

The Court of Appeal in White v Alder [2025] confirmed that informal boundary agreements can bind future owners.
In this article, John Aldis explains the difference between boundary creation and demarcation agreements and their role in dispute resolution.

1. The Court of Appeal recently confirmed in White v Alder [2025] EWCA Civ 392 that: 
   1. A boundary agreement will bind successors in title regardless of whether they purchased on notice of the agreement; and
      
   2. Boundary agreements do not need to comply with s.2 of the Law of Property (Miscellaneous Provisions) Act 1989 – and indeed may be oral or implied from conduct – so long as they entail the transfer of no more than a “trivial amount of land”.
2. In doing so, the court considered five other Court of Appeal authorities which have approved the seminal case, Neilson v Poole (1969) 20 P&CR 909, and confirmed that it should be preferred to the most recent decision, Gibson v New [2021] EWHC 1811 (QB), in which Murray J expressed the obiter view that a boundary agreement “has no proprietary effect between third parties”.
3. This raises the following questions:
   1. What is a boundary agreement and why is it binding on successors in title?
   2. What weight will the courts give to a boundary agreement when determining a boundary?
   3. How should boundary agreements be used in dispute resolution?

WHAT IS A BOUNDARY AGREEMENT AND WHY IS IT BINDING SUCCESSORS IN TITLE?

4. Asplin LJ (with whom Zacaroli LJ and Sir Launcelot Henderson agreed) answered these questions so concisely in White v Alder [2025] EWCA Civ 392 that the following paragraphs are likely to be frequently copied and pasted. She said (annotations in square brackets added):
   
   53. “Where does that leave us? Drawing all of the authorities together, it seems to me that as Megarry J first explained [in Neilson v Poole], there are two types of boundary agreement. The first is an agreement the purpose of which is to move a boundary so as to transfer land from one neighbour to another. That first type is subject to the formalities necessary for the transfer of land. The second type is an agreement, the purpose of which is to define a previously unclear or uncertain boundary, even if it includes the conscious or unconscious transfer of a trivial amount of land. It is presumed that the land transferred is trivial unless the presumption is rebutted. This second type, a boundary demarcation agreement, binds the parties to it for the reasons explained by Briggs LJ in Nata Lee Ltd [Nata Lee Ltd v Abid [2014] EWCA Civ 1652] at [32]. The consideration for the agreement is the substitution of certainty for uncertainty and the avoidance of the risk of future disputes. Although in many cases, the parties will act upon the agreement, for example, by building a wall or erecting a fence, there is no need for anything more in order to render it binding as between them. That is clear, not only from Neilson v Poole itself which has been consistently endorsed in the Court of Appeal, but also from Stephenson  [Stephenson v Johnson [2000] EG 92 (CS)] where it was stated that it was unnecessary to consider the estoppel argument in addition to the implied boundary demarcation agreement, from the clear obiter dicta of Lawrence Collins LJ in Haycocks  [Haycocks & Anr v Neville & Anr [2007] EWCA Civ 78] and the approach adopted by Briggs LJ in Nata Lee Ltd.
   54. Such an agreement has proprietary effect and, as a result, also binds successors in title. It does so because of its very nature. It defines and delineates the boundary between the properties as from the root conveyance or transfer. Such an agreement is, of its very nature, a delineation of the property transferred or conveyed and is so for all purposes. As no one is able to transfer or convey more than they own, such an agreement effectively “binds” successors in title whether or not they have knowledge of it. It does so because it defines what they purchase. As “Ruoff and Roper on the Law and Practice of Registered Conveyancing” explains at paragraph 5.020, a legal boundary does not move because the land is subsequently conveyed or transferred. The boundary demarcation agreement is ancillary to the conveyance or transfer. As Megarry J explained at 919 of Neilson v Poole , the boundaries established are, in the words of Lord Hardwick L.C. in Penn v Lord Baltimore [Penn v Lord Baltimore (1750) 1Ves. Sen. 444], “presumed to be the true and ancient limits”. In other words, a boundary demarcation agreement establishes on the ground the physical extent of the respective legal estates created by the conveyance or transfer. The boundary is presumed always to have been in that location.
   55. In the case of unregistered land, a vendor cannot convey to a purchaser more land than he owns. In the case of registered land, upon registration the purchaser is deemed to be the proprietor of the registered estate: section 58 of the Land Registration Act 2002 (“LRA 2002”) . The boundary of that registered estate will, however, only be a general boundary unless exceptionally it has been determined under section 60(1) LRA 2002. A general boundary does not determine the exact line of the boundary: section 60(2) LRA 2002 . The exact boundary of registered land may, therefore, be established by a boundary demarcation agreement.
   56. In either case, if the extent of the land owned by the vendor has been conclusively established by a boundary demarcation agreement, the purchaser cannot acquire title to more land than was owned by the vendor. If the effect of a boundary agreement is to define the extent of the parcels of land owned by the parties to it, it must logically continue to define the extent of those parcels when they are transferred to a successor in title. 
   57. Although a boundary demarcation agreement which is implied may be more difficult to prove than an express agreement, there is nothing to suggest that the underlying principle should be different. In just the same way, the agreement defines the parcel of land and as a result, defines what is transferred to a successor in title. This is consistent with the decisions in Burns and Stephenson where the boundary agreement which bound successors in title was implied from the conduct of predecessors.”
5. Evidently, there is a wide spectrum between:
   1. The first type of boundary agreement (which I will refer to as a “boundary creation agreement”); and
   2. The second type (which Asplin LJ defines as a “boundary demarcation agreement”).
6. A boundary demarcation agreement which is inferred from conduct may look more like a conversation than a contract. For instance, in Acco Properties Limited v Severn [2011] EWHC 1362 (Ch), the defendant and the claimant’s predecessor in title had a discussion about trees in which the latter said: “They are on your land. … Take the trees down at your own expense”. HHJ Barker held that this amounted to an oral boundary demarcation agreement. He said:
   66. “…I am satisfied, by reference to the oral evidence of Mr Parker, Mr Severn and Mrs Severn, that when they met in March 2008 there was a sufficient discussion of the boundary for the defendants to understand that when cutting down the two trees…they would be cutting down trees on their own land, and thus within their boundary.”
7. (As an aside, the judge also found that the defendant’s alleged written boundary agreement was a forgery, noting that his credibility was undermined by the fact that he had an SAS tattoo despite never having served in the special forces).
8. This raises the question of what weight should be given to the various types of boundary agreement, especially when they are hard to distinguish from the other extrinsic evidence which may be relevant. 

WHAT WEIGHT WILL THE COURTS GIVE TO A BOUNDARY AGREEMENT WHEN DETERMINING A BOUNDARY?

9. The key principles for determining boundary lines were stated by HHJ Barker QC in Acco Properties Limited v Severn as follows: 
   11. “Before turning to the facts and expert opinion evidence, I should briefly remind myself of the principles relevant to determination of boundary disputes, at least insofar as they are potentially applicable to this case:
       1. Where, as in this case, the property in question is registered land, the file plans show only general boundaries and not the exact line of the boundaries unless the property is said to be “more particularly described in the plan.”
       2. Similarly, Ordnance Survey plans, if not forming part of the registered title as filed plans, are no more than a general guide to a boundary feature, and they should not be scaled up to delineate an exact boundary. This is because the lines marking the boundaries become so thick on being scaled up as to render them useless for detailed definition.
       3. In order to determine the exact line of a boundary, the starting point is the language of the conveyance aided, where the verbal description does not suffice, by the representation of the boundaries on any plan, or guided by the plan if that is intended to be definitive.
       4. If that does not bring clarity, or the clarity necessary to define a boundary, recourse may then be had to extrinsic evidence – such as topographical features on the land that existed, or maybe supposed to have existed, when the dividing conveyance was executed.
       5. Admissible extrinsic evidence may also include evidence of subsequent conduct where of probative value in showing what the original parties intended.
       6. Evidence of later features – that is, later than the earliest dividing conveyance – may or may not be of relevance. The probative significance of such evidence depends upon the extent to which, if at all, the dividing conveyance, or evidence of its terms, exists.
       7. Where a boundary is in dispute, it is important to bring certainty to the determination by proclaiming the boundary and not leaving the plot “fuzzy at the edges” (Neilson v Poole (1969) 20 P&CR 909, Megarry J).
       8. Even where a boundary line may be determined by reference to a conveyance, other evidence may be admitted and probative in establishing a different boundary obtained by adverse possession, showing enclosure of the land in denial of the title of the true owner. As the phrase implies, title is established by intentionally taking exclusive possession of land without the consent of, and adverse to the interests of, the true owner, and maintaining such possession continuously for the limitation period.
       9. As to informal boundary agreements, the statutory requirement that contracts for the sale or other disposition of land be in writing does not apply. That is because the purpose of such agreements is to demarcate an unclear boundary referred to in title documents and not to transfer an interest in land.
       10. Such agreements are usually oral and the result of neighbours meeting to avoid or resolve a potential or actual dispute. However, there is scope for a boundary agreement to be implied or inferred – that is, to be the logical conclusion to be drawn from primary facts.
       11. When bearing these principles in mind as the platform on which to place and examine the facts, a judge should have regard to three further important yardsticks or rules of thumb. These are: (1) when considering any acquisition of property, it is vital to consider what a reasonable layman would think he was buying; (2) every case turns on its own facts; and (3) the task of the court is to assess all available and admissible material in arriving at its answer, and then to achieve the correct answer.”
10. Accordingly, this would appear to be the correct order in which to consider any boundary agreements (to the extent that different types of evidence can be compartmentalised, and ignoring adverse possession):

Stage 1

1. The starting point is the language of the transfer deed which created the boundary (“the dividing transfer”) aided, where the verbal description does not suffice, by the representation of the boundary on any plan (especially if it is intended to be definitive). 
2. A s.2 compliant boundary creation agreement should presumably be considered at Stage 1. It may have been completed by a registered TP1 transfer. If so, the TP1 will be the new dividing transfer, and it (rather than the agreement) will be the starting point for any boundary determination. If not, the agreement will give rise to a specifically performable obligation to transfer (see Neilson v Poole at 918-919).
3. However, the TP1 (or boundary creation agreement) will only be the end point if the boundary is defined (either by words or on a plan) with sufficient particularity. This is not the case with most dividing transfers – which is why the court often needs to introduce extrinsic evidence to interpret where the original parties intended the boundary to be. 

Stage 2

4. If the dividing transfer is not sufficiently clear to define the boundary, recourse may then be had to extrinsic evidence which is of probative value in showing what the original parties to the dividing transfer intended (such as topographical and boundary features which existed at, or shortly after, the dividing transfer). This evidence often includes historic photos, satellite images, old OS plans, half-buried bricks and rotting fence posts. 

Stage 3

5. The court may then examine any subsequent discussions and conduct to determine whether there is boundary demarcation agreement. This is evidence that the original parties, or their successors in title, had a meeting of minds about the correct interpretation of the dividing transfer. 
6. A boundary demarcation agreement will not be the starting point for determining the boundary because the court must first seek to determine the original location of the boundary and then consider whether the presumption that the boundary demarcation agreement entails no more than a trivial transfer of land has been rebutted. If the presumption is rebutted, the agreement is in fact a boundary creation agreement and it will need to comply with s.2 to be enforceable. 
7. However, Stage 3 may merge into Stage 2 if the evidence for the meeting of minds is the agreed placement of a boundary feature by the original parties to the dividing transfer shortly after it was completed, because that would be strong evidence of their original objective intentions.
8. It will often ultimately be the case that informal boundary demarcation agreements are the end point of the court’s determination because they generally provide a higher level of specificity about the location of the boundary than the dividing transfer. For example, in Acco Properties Limited v Severn, the dividing transfer could not be located, and the extrinsic evidence was inconclusive. But HHJ Barker QC was able to use the informal boundary demarcation agreement to determine the boundary. After identifying the agreement mentioned above, he said: 
   66. “… From that, their boundary could be plotted as a straight line to the north of the tree trunk…shown at the photograph on page 295…That boundary line would, in my judgment, be the result of an informal boundary agreement to demarcate the boundary and therefore take precedence over (in so far as different from) the boundary line that I had found by reference to the extrinsic evidence.”

11. In summary, a boundary creation agreement may be the appropriate starting point for a boundary determination. But if it (or the associated TP1) does not define the boundary with sufficient precision, it will not be the end point. Whereas an informal boundary demarcation agreement will often be the end point (but not the starting point) of a boundary determination – because it will often relate to a specific issue, such as whether a fence should be 3cm further east or west.

HOW SHOULD BOUNDARY AGREEMENTS BE USED IN DISPUTE RESOLUTION?

12. Since boundary agreements are so helpful in clarifying boundaries, one might assume that the best way to avoid boundary disputes would be to require the undetermined boundaries of any plot of land to be agreed when the plot is next sold. However, in Clapham v Narga [2024] EWCA 1388, Newey LJ explains that this experiment was tried in the Land Registry Act 1862, but promptly abandoned because (as the Land Transfer Commission commented in 1870):
    1. “…people served with notices immediately begin to consider whether some injury is not about to be inflicted on them. In all cases of undefined boundary they find that such is the case, and a dispute is thus forced upon neighbours who only desire to remain at peace”. 
13. That is why the general boundary on a title plan only identifies the property being bought, rather than its exact boundaries. It is better not to give your prospective neighbour an opportunity to consider whether the fence is 3cm out of line if no-one has ever raised the point before. 
14. However, if neighbours are already in a boundary dispute, or if a small boundary discrepancy is discovered during pre-contractual enquiries for the purchase of land, a boundary demarcation agreement is a cheap and effective way of clarifying (and, perhaps, partially re-drawing the boundary). It is certainly easier than completing a TP1 transfer of part of one title and then either registering a new title for that part or applying to amalgamate it with the existing title. 
15. Incidentally, this is what happened when my wife and I bought a house last year. Our conveyancer required the vendor to procure a TP1 transfer of a 1m² section of a loo which has been protruding into the neighbouring property for nearly 200 years. I proposed a boundary demarcation agreement, but the conveyancer doubted whether it would give good enough title to satisfy the mortgagee. So, we now have one of the only loos in the country with its own flying freehold title. 
16. In any case, when a boundary demarcation agreement is appropriate, it should be carefully drafted to ensure it puts all future disputes to rest. In Practice guide 40: HM Land Registry plans, supplement 4, boundary agreements and determined boundaries, HM Land Registry gives practical guidance on drafting effective boundary demarcation agreements which can either be noted on the registers of title or form the basis of an agreed application for a determined boundary under s.60 LRA 2002.  They state this is in relation to plans:
    1. “2.2…It would be sensible for both parties, and any witnesses, to sign and date the plan attached to the agreement. Generally speaking, the better the quality of the plan – and, in particular, the more precisely it shows the position of the legal boundary – the more helpful the agreement is likely to be to the parties and their successors in title. You may, therefore, wish to ensure that the plan complies with the guidelines in practice guide 40 supplement 2: guidance for preparing plans for HM Land Registry applications. But there will be cases where the particular circumstances mean that a more basic plan is sufficient.
    2. This is not to suggest that a plan of some sort will always be necessary. In some cases, it may be thought enough to identify the boundary in words alone. For example, the agreement might include a clause along the following lines:

“The parties agree that the legal boundary between the land within their respective registered titles is the centre line of the wall running between their properties.”

17. In conclusion, White v Alder confirms that the policy of the courts and HM Land Registry is to encourage neighbours to resolve or prevent boundary disputes for themselves and all successors in title with the use of cheap and effective boundary demarcation agreements. I would certainly be happy to draft more boundary agreements and fewer skeleton arguments. 

Whilst every effort has been taken to ensure that the law in this article is correct, it is intended to give a general overview of the law for educational and/or informational purposes. It is not intended to be a substitute for specific legal advice and should not be relied upon for this purpose.

This article represents the opinion of the author and does not necessarily reflect the view of any other member of St Philips Chambers.