Siege Law


| Mar 4, 2022


Images of the million or more persons who have fled toward neighboring States highlight the tragic humanitarian impact of the ongoing armed conflict between Ukraine and Russia. However, a worse fate could await those unable to flee. Despite a slower than expected advance, Russian forces appear to be maneuvering to lay siege to several Ukrainian cities. While reports indicate the port city of Mariupol is  already under siege, cut off from critical supplies and support.

The term siege refers generally to a military effort to surround and cut off an area, often but not always a city, to deny external access or egress, and secure the defender’s submission by deprivation or isolation.

Because of their devastating human costs, sieges have inspired specific law of war rules and legal considerations. But as with legal limits on war generally, these rules reflect a compromise between human needs and military demands. Siege rules fully vindicate neither humanity nor military necessity; each concedes something to the other. The current conflict may well reveal whether these rules, and the compromises they reflect, are relevant, practical, and optimal in both military and humanitarian terms.

Why Siege?

Siege—or encirclement as military doctrine refers to it—is an essential aspect of modern military operations. Armies resort to the tactic for a variety of reasons. But chief among them is to avoid slow, costly, and bloody urban ground assaults. Military planners usually estimate that significant numerical advantage is required for successful assaults. Encircling and waiting out an enemy, rather than attacking prepared positions, can lower those numerical demands significantly, reducing losses and freeing forces, weapons, and supplies for other efforts.

Historically, isolation has been the key to effective siege; indeed, it’s probably its sine qua non. History is packed with examples of imperfectly isolated sieges that have failed, including at Vienna, Leningrad, Sarajevo, and Grozny just to name a few roughly Eastern European examples (The Mongol Empire successfully sieged and sacked Kyiv by assault in 1240). Even slight cracks in the besieging force’s effort to seal off its target can tip the balance in favor of the besieged force, buoying morale and will to resist.

It’s unsurprising then that military doctrine instructs planners and commanders to achieve complete isolation early and to maintain it throughout a siege. Current U.S. operational doctrine speaks of physical, psychological, and most recently electronic isolation. Each form of isolation is essential to achieving the enemy attrition that in assaults is achieved by direct attack. Simply put, isolation is paramount to effective siege.

Means and Methods Law and Siege

The law of war treaties and customs that regulate the means and methods of attacks apply fully to siege operations. These rules have matured into an elaborate collection of duties and precautions too prolific to fully address here. But the Regulations annexed to both the 1899 Hague Convention II and 1907 Hague Convention IV, adopted a rule specific to siege. It outlines duties applicable to both besieging and besieged forces. Article 27 of the Hague Regulations reads:

In sieges and bombardments all necessary steps must be taken to spare, as far as possible, buildings dedicated to religion, art, science, or charitable purposes, historic monuments, hospitals, and places where the sick and wounded are collected, provided they are not being used at the time for military purposes.

It is the duty of the besieged to indicate the presence of such buildings or places by distinctive and visible signs, which shall be notified to the enemy beforehand.

The article’s extensive list of protected objects is eye-catching. Less prominent, but just as important however, are its caveats. The important facilities the article protects are only spared until they are used for military purposes. In practice, the almost inevitable intermingling of city defenders and besieged forces with their urban surroundings and their need to use everything at their disposal to resist cut deep into the article’s protective scheme. Subsequent rules have elaborated on defenders’ duties, as explained in this symposium contribution by Professor Eric Jensen, to separate their forces from civilians and civilian objects. But the congested context of siege invariably limits defenders’ options in this respect.

The literal phrasing of the Article 27 obligation to spare is promising in humanitarian terms. However, it is likely that interpretation of the phrase “as far as possible” has been tamed in practice to mean “as far as feasible.” In law of war terms, feasibility increasingly refers to what can be done in very practical terms considering a wide range of military and humanitarian factors including competing demands on time, people, and other resources. What is “possible” is not likely to include everything that can be done in light of the full economic and military potential of the State. Rather it refers only to what can be done with what is immediately on hand and consistently with both the overall and immediate goals of the operation.

Spotty application and qualified practice aren’t limited to the besieging force’s Article 27 obligations. The defender’s duties to visibly identify protected objects does not seem to have been widely practiced either. Other treaties, including the 1954 Hague Convention for the Protection of Cultural Property have enhanced and somewhat clarified that obligation but with similarly limited practice or real implementation in combat. Technology, however, may offer parties greater potential to communicate protected sites through other means. This potential may be tested in the current conflict by brave humanitarian intermediaries.

Though it’s not dedicated exclusively to sieges, a law of war rule that addresses starvation has special relevance to them. Article 54(1) of Additional Protocol I to the 1949 Geneva Conventions, to which both Ukraine and Russia are parties, provides, “Starvation of civilians as a method of warfare is prohibited.” That seemingly simple rule finds support elsewhere including likely status as a customary prohibition applicable to all States and in a more recent expression of war crimes law. States included the civilian starvation prohibition in the war crimes jurisdiction of the International Criminal Court. Neither Ukraine nor Russia is a party to the Court’s statute, however in 2014 and 2015 Ukraine filed declarations committing its situation to the Court’s jurisdiction.

Of course, the devil of the starvation rule is in its details and interpretation. Frankly, it’s unclear how the imposed deprivations at the heart of siege fare under this rule. By one interpretation, the rule prohibits any starvation of civilians, including as an incidental effect of an effort to starve defending forces into submission. But another interpretation maintains that only starvation directed specifically at civilians is prohibited. By this view incidental though foreseeable effects of civilian starvation are not prohibited, although they must not be disproportionate, that is, excessive in relation to anticipated military advantage. Although this latter view is held by the United States (§. 5.20), it reduces the rule’s humanitarian effect, perhaps to the vanishing point.

Still, the military implications of this debate may be enormous. The former expansive interpretation of the starvation rule could well render siege impossible as it has historically been known. The steps required in terms of evacuation, or admitting relief to the besieged area that will almost certainly find its way to the besieged force, render isolation practically impossible. For instance, the U.S. Department of Defense only repealed in 2015 legal guidance that permitted besieging forces to fire on persons escaping a besieged area. The former guidance, no doubt in appreciation of the isolation imperative, had advised, “Persons who attempt to leave or enter a besieged place without obtaining the necessary permission are liable to be fired upon, sent back, or detained.” (para. 44a.)

No rule applicable to siege perhaps illustrates in starker terms the compromise required between military and humanitarian interests.

Humanitarian Relief Law during Siege

Three further law of war rules are also either directed specifically to or are especially relevant to sieges.

First, Article 17 of the Fourth Geneva Convention states,

The Parties to a conflict shall endeavour to conclude local agreements for the removal from besieged or encircled areas, of wounded, such, inform and aged persons, children and maternity cases, and for the passage of ministers of all religions, medical personnel and medical equipment on their way to such areas.

While a promising commitment and addressed to the most vulnerable persons in siege, the rule’s caveats of “endeavour” and “agreements” greatly limit its legal and practical import. By its plain meaning the article’s legal obligation is only to try or make attempts at agreements. And both the besieging force and the besieged force must have agreed for the evacuations and passages of relief to be compulsory. Additionally, substantial classes of persons not named in the article do not benefit from its obligation. Otherwise healthy civilian adults may be forced to stay in the besieged area by either the besieging or besieged force.

Second, Article 23, also of the Fourth Geneva Convention, compliments the Article 17 limited evacuation and relief passage agreement scheme. It is lengthy and somewhat convoluted but essentially requires that parties to a conflict allow passage of a limited class of relief supplies for civilians, and more for particularly vulnerable classes of persons. But, not surprisingly, it operates only if the parties are satisfied no advantage will result “to the military efforts or economy of the enemy ….”

Last, Article 70 of an Additional Protocol that updates the Geneva Conventions, including the Fourth Convention, addresses humanitarian relief actions. Mitigating the limits of preceding rules, it requires that parties admit a wider range of relief supplies. Although it prioritizes vulnerable groups, it also speaks to civilians more broadly. Still, the Article does not offer an unlimited guarantee of relief. Its obligations operate, again unsurprisingly, “subject to the agreement of the Parties concerned ….”

The military imperative of isolation looms large over the chances that any of the above obligations will take effect. Surely the need for isolation explains the limited reach and reduced humanitarian promise of siege rules. An important and increasingly influential United Nations-sponsored academic product, the Oxford Guidance on the Law Relating to Humanitarian Relief Operations, lends more bite to these obligations. It understands the agreement elements of each rule in a more compulsory sense, offering significant humanitarian potential. Yet it’s still unclear how well that study reflects the legal opinions of States and how well it accounts for the isolation imperative, a question I’ve addressed elsewhere.

In, for now, an encouraging sign, Ukraine and Russia have reportedly agreed to organize safe corridors to evacuate civilians and deliver humanitarian supplies. The extent to which these agreements will be observed or hold as pressure mounts to maintain isolation of besieged cities will be important not only to the fate of war’s victims but also to the law.


As a reactive regime, the law of war has made progress regulating sieges. For most of historical experience the customs of war tolerated merciless looting and the deliberate murder of defeated military personnel and civilians alike following sieges. Fortunately, siege law has abandoned these customs, and adoption of the rules highlighted in this post have softened siege, if only slightly and sporadically. Nonetheless, the legal truth, difficult for many to accept, is that a harsh legal regime applies to sieges. It is law best understood, interpreted, and applied through military operational realities, especially the isolation imperative.

But it is also law that may be out-of-touch with prevailing sensibilities. While not determinative, civil society’s reactions to the suffering involved in sieges, including those currently or soon to be underway in Ukraine, may inspire a reexamination of siege law by States. But meanwhile, fidelity to that law as we find it, rather than as we wish it was, seems best for legal discourse and advice, leaving political and policy-based improvements to their respective fora.

In that vein, on 31 March and 1 April the West Point Lieber Institute, the Harvard Law School Program on International Law and Armed Conflict, and the Washington D.C. Regional Delegation of the International Committee of the Red Cross will co-host a workshop on Siege and the Law of War. We will convene not only some of the world’s leading academic legal minds and humanitarian professionals but also an impressive contingent of military commanders, planners, and experts. We hope this truly pluralistic approach will reveal helpful insights into the sadly but truly relevant content of siege law.


Sean Watts is a Professor in the Department of Law at the United States Military Academy, Co-Director of the the Lieber Institute for Law and Land Warfare at West Point, and Co-Editor-in-Chief of Articles of War.