A Blockade by the Book: Why CENTCOM’s Carve-Out for Non-Iranian Ports Matters

[Frederik Rogiers is a PhD researcher and teaching assistant at the Ghent Rolin-Jaequemyns International Law Institute and Ghent Maritime Institute, Faculty of Law and Criminology, Ghent University]

On 12 April 2026, following the collapse of the Islamabad talks, President Donald Trump announced on Truth Social that the United States Navy would impose a “naval blockade of the Strait of Hormuz”. Within twenty-four hours, U.S. Central Command (CENTCOM) restated the operation in markedly narrower terms: its press release announced that:

“the blockade will be enforced impartially against vessels of all nations entering or departing Iranian ports and coastal areas”

while “CENTCOM forces will not impede freedom of navigation for vessels transiting the Strait of Hormuz to and from non-Iranian ports.” On 16 April, General Dan Caine made the distinction even more explicit: “The U.S. action is a blockade of Iran’s ports and coastline, not a blockade of the Strait of Hormuz.” The corresponding Notice to Mariners confirmed that enforcement would be restricted to the Gulf of Oman and the Arabian Sea, east of the Strait itself.

This post argues that the shift in wording between the President and CENTCOM was not merely cosmetic but legally decisive: the geographical carve-out for traffic “to and from non-Iranian ports” is the juridical keystone on which the lawfulness of the entire operation rests. Without it, the blockade would collapse into an unlawful belligerent closure of a neutral international strait. With it, the operation conforms to the conditions of a lawful blockade as these have been developed from the Paris Declaration of 1856 through the San Remo Manual and the Newport Manual.

What this post does, and does not, address

The analysis below is deliberately narrow. It concerns the U.S. naval blockade of Iranian ports as that blockade was declared on 12–13 April 2026. It does not assess: (i) Iran’s closure of the Strait of Hormuz, the legality of which has already been examined in depth by Konstantinos Deligiannis-Virvos on this blog, by Alexander Lott on EJIL:Talk! and by Mark Nevitt on Just Security; (ii) the CENTCOM update of 16 April purporting to extend belligerent visit-and-search authority to all Iranian vessels and vessels suspected of carrying contraband regardless of geographic location; (iii) attacks on merchant vessels, on which see Schmitt and McLaughlin’s Q&A; (iv) questions of prize, contraband, convoy, and the belligerent right of visit and search on which see e.g. Bamnios’ post on this blog; or (v) the jus ad bellum legality of the American–Israeli strikes of 28 February. Several of those issues are treated in the comprehensive Just Security collection on the conflict.

Iran’s broader characterisation of CENTCOM’s operation as “piracy and maritime theft under the guise of a blockade” is rhetorically convenient for Iran but legally unacceptable. Articles 101 and 102 UNCLOS by design exclude state conduct by warships under command authority (see Douglas Guilfoyle, Shipping Interdiction and the Law of the Sea, ch. 4). Furthermore, the 1988 SUA Convention (Art. 2(1)) together with its 2005 Protocol (new Art. 2bis) confirm that warships and armed-forces activities during armed conflict fall outside their scope. The relevant body of law is the customary law of naval warfare, applied alongside UNCLOS.

The CENTCOM Carve-Out and the Five Conditions of Lawful Blockade

The existence of an international armed conflict between the United States and Iran is, at this stage, beyond serious dispute: the Tadić threshold of “a  resort to armed force between States” – is plainly met. The conditions for a lawful blockade are likewise reasonably settled, even if their most authoritative articulation lies in instruments that are themselves non-binding, i.e.  the 1994 San Remo Manual (paras. 93–104) and  the 2025 Newport Manual (chapter 7.4). Both qualify as influential expert restatements, yet – as the divergence over proportionality below illustrates – are not entirely uniform on the content of the customary law they purport to capture. These conditions and their divergent interpretations have also been developed in literature – most exhaustively in Phillip Drew’s 2017 The Law of Maritime Blockade: Past, Present, and Future, and most accessibly in Heintschel von Heinegg’s Max Planck entry.

Declaration and notification

First, in order to be lawful, a blockade must be declared and notified to belligerents and neutrals alike (Rule 93 San Remo Manual; para. 7.4.2 Newport Manual). The declaration must contain the date and time of commencement, the blockade’s duration, its location and extent, and the grace period offered to neutral vessels in port to depart (Rule 94 San Remo Manual). CENTCOM’s press release of 12 April, with entry into force at 10:00 ET on 13 April, satisfied each element: it specified the geographical extent (Iranian ports on the Arabian Gulf and Gulf of Oman), the commencement date, and the carve-out for traffic transiting the Strait to and from non-Iranian ports, and expressly provided that neutral vessels already in Iranian ports could depart during a twenty-four-hour grace period. A formal Notice to Mariners was published the same day. On this point the operation is, as Huntley puts it, “serious but not literal”: the official act of the blockading State, and not the Truth Social post that preceded it, is what enters the legal record.

Effectiveness

The second condition – that a blockade must be “effective” – was historically designed to extinguish the practice of “paper blockades” (paragraph 4 of the 1856 Paris Declaration). The unratified but doctrinally influential 1909 London Declaration (Art. 3) added that effectiveness is “a question of fact”. Modern doctrine permits the blockading force to be “stationed at a distance determined by military requirements” (Rule 96 San Remo Manual). The Newport Manual summarises the contemporary test as a “sufficient degree of probability of capture”, expressly rejecting any 100% standard. Heintschel von Heinegg, (para. 33), formulates the test as a:

“high probability that ingress to and egress from the blockaded area will be detected, and prevented by the blockading power”

(see also para. 13.10.2.3 US Law of War Manual).

Enforcing the blockade in the Gulf of Oman and Arabian Sea rather than in the Strait itself or directly off the Iranian coast is, following the above, plainly lawful as a matter of operational geography.

The factual picture on the water is more complicated. By 20 April, Lloyd’s List had counted at least 26 shadow-fleet vessels crossing in both directions, with other sources going as high as 34. Many of these vessels, however, seemingly only crossed the Strait of Hormuz – which the blockade does not close – rather than the blockade line a few hundred miles seaward in the Gulf of Oman. The Pentagon later denied Lloyd’s report, supported by reports from JPMorgan and other sources now confirming that neither tankers nor seaborne crude had left the blockaded area for a number of weeks. Indeed, CENTCOM’s latest figures (1 June) show that so far 121 vessels have been redirected, five disabled and three seized, including the Iranian-flagged Touska and Hasna, while the US struck at the Gambian-flagged Lian Star on 30 May and the Botswana-flagged M/T Lexie on June 2. Mid-May, Adm. Brad Cooper claimed the operation permitted “zero trade” to or from Iran.

It’s important to recall – as stated above – that effectiveness never required impermeability, only a force sufficient to make ingress or egress dangerous. In other words, an occasional breach does not defeat the entire blockade. What the law asks is a real risk of capture, not an unbroken record. It is the standard on which the much-run Union blockade was upheld in the Prize Cases, and on which the 2011 UN Palmer Report found Israel’s Gaza blockade lawful despite repeated flotilla attempts to breach it. This risk of capture certainly seems present in the case at hand, evidenced by the above numbers and examples. Finally, one should recall that as the blockade went on and enforcement tightened, no further reports of successful blockade breaches came out, while much of the original “bypassing” consisted of ship-to-ship transfers and cargo-laundering the United States does not legally treat as a breach of blockade. In total, the forces now deployed – over 15,000 personnel, two carrier strike groups and 200-plus aircraft and warships – plainly generate the probability the law requires. The point stays dynamic, however: were port-running to increase and become systematic and sustained, effectiveness and lawfulness would fall to be reassessed; but the trend so far has run toward tighter, not looser, enforcement.

Impartiality

Third, a blockade must be applied impartially to the vessels of all States (Rule 100 San Remo Manual; para. 7.4.4 Newport Manual). The condition is older than San Remo: it is recognisable in Article 5 of the unratified but doctrinally influential 1909 London Declaration, and indeed reaches back to Paragraph 4 of the 1856 Paris Declaration in its concern with neutral commerce. General Caine’s 16 April restatement reaffirmed that the blockade applies “to vessels of all nations”.

While CENTCOM has not released a comprehensive flag-state breakdown of the redirected vessels, the named enforcement actions already span several registries — Iranian-flagged, stateless and flag-of-convenience alike. The harder impartiality question lies precisely with that last category. China is the largest destination for Iranian crude by far, but the trade is overwhelmingly conducted through flag-of-convenience vessels with opaque ownership chains – the Chinese-owned, Malawi-flagged MV Rich Starry is one publicly-reported example. The structural test for impartiality is therefore not whether Chinese-flagged vessels have been interdicted (very few exist on the relevant routes) but whether vessels objectively comparable to the Touska are being waved through or interdicted based on who ultimately benefits from their cargo. Framed doctrinally rather than as a matter of optics, such selective non-enforcement would breach Rule 100’s requirement that the blockade bear equally on the vessels of all States, while importing a judgment about a cargo’s enemy destination that belongs not to the law of blockade, but to the separate law of contraband – the very regime under which CENTCOM expanded visit-and-search on 16 April.  On the public record to date no such selectivity is visible. The named enforcement actions span Iranian, stateless and third-flag vessels alike, regardless of destination. As such – and for now – one may conclude that the requirement of impartiality remains met, even if the absence of a published flag-state breakdown keeps that assessment provisional.

Humanitarian limits and the proportionality question

A blockade is furthermore prohibited if “the sole purpose [is] to starve the civilian population or to deny it other objects essential for its survival”, or if damage to the civilian population is “excessive in relation to the concrete and direct military advantage anticipated” (Rule 102 San Remo Manual). Rule 103 obliges the blockading party to allow free passage of essential foodstuffs and medical supplies when the civilian population is “inadequately provided”, and Rule 104 imposes safeguards on inspection.

The two parameters at issue here – sole purpose and proportionality – are doctrinally contested. The Newport Manual departs from the San Remo Manual by rejecting the proportionality rule as applicable to blockade, reasoning that a blockade is not an “attack” in the technical sense and that no consistent State practice supports applying proportionality to blockade. Interestingly, DoD’s Law of War Manual (para. 13.10.2.5) takes the contrary – San Remo – view. The literature is similarly divided: in his EJIL: Talk! discussion of the Mavi Marmara Flotilla Report, Yuval Shany has defended applying the San Remo Rule 102 proportionality test to naval blockade – an approach endorsed by the Palmer Report itself; Eyal Benvenisti has gone further in his study of the Gaza blockade, urging that the “full scope of damage … suffered by the civilian population” be weighed against the military purpose. The opposing view, defended within the Newport drafting committee, is that proportionality stricto sensu is not the right doctrinal tool for an instrument that operates by economic, not kinetic, means, while accepting that:

“a denial of the blockading power’s consent to relief consignments will be arbitrary only in exceptional circumstances in which the very survival of the civilian population is at stake …”

As Wolff Heintschel von Heinegg’s puts it in his analysis on the Lieber Institute site: “paragraph 102(b) of the San Remo Manual … is not accepted as reflective of customary international law”.

The choice of which Manual is more persuasive need not be resolved here. For present purposes, it is enough to observe that the blockade pursues no humanitarian end but an avowedly economic one as also stated by Defense Secretary Hegseth, with the purposes being to throttle the Iranian Regime’s oil revenue, not deprive the civilian population of the necessities of life. Furthermore, the CENTCOM Notice to Mariners expressly permits humanitarian shipments subject to inspection. As such, the bar on starvation as a blockade’s sole purpose is quite certainly not engaged and even on San Remo’s more demanding threshold the operation passes, for it is – as of yet – not excessive in its effects as Iran still also retains extensive landward supply routes via its neighbours. The picture is, however, dynamic. Should demonstrable civilian suffering of the kind documented in Gaza emerge, the proportionality assessment will have to be revisited regardless of where one sits on the San Remo/Newport divide.

Access to neutral ports and coasts – where the carve-out does the work

Finally, there is the question of the access to neutral ports and coasts, the condition the carve-out exists to satisfy, and it is here that the substitution of words between 12 and 16 April carries its full juridical weight. Rule 99 of the San Remo Manual provides that “[t]he establishment of a blockade is prohibited if … it has the effect of barring access to neutral ports or coasts” (see also para. 7.4.6 Newport Manual). On the Arabian littoral of the Persian Gulf alone sit the major ports of six neutral States: the United Arab Emirates, Saudi Arabia, Qatar, Bahrain, Kuwait and Iraq. Together they handle a substantial share of the roughly one-fifth of global oil consumption that transits the Strait. Trump’s 12 April post, read literally, would have barred neutral access to all of them; the operation it described would have failed Rule 99 on its face. The CENTCOM substitution – enforcement against vessels “entering or departing Iranian ports”, with explicit preservation of “freedom of navigation for vessels transiting the Strait of Hormuz to and from non-Iranian ports” – moves the operation from the prohibited category into the permitted one.

This is not a technicality. Rule 99 reflects one of the most consistent threads running through the law of blockade since its inception: a belligerent may close the enemy’s coastline, but it may not interpose itself between neutrals and their own ports. The carve-out is, in this sense, the operational implementation of Rule 99 applied to the unusual geography of a coastline that forms one side of an international strait – which the US can blockade vis-à-vis Iran only. It is also the line that distinguishes lawful belligerent closure of enemy ports from unlawful belligerent closure of an international strait, the latter being precisely the conduct that has been attributed to Iran since March (and analysed at length in the Deligiannis-Virvos post cited above).

It is worth mentioning that, as Nevitt observes, the blockade does not, and cannot, apply in neutral territorial seas: vessels skirting the Omani or Pakistani coastline, invoking the right of innocent passage cannot be intercepted there. While Oman or Pakistan systematically permitting blockade-running through their territorial seas raises neutrality concerns of its own, those concerns lie outside the legality of the U.S. blockade as such.

Concluding Remarks

The chronology of CENTCOM’s operation – Trump’s Truth Social post on 12 April, the press release within twenty-four hours, and Caine’s restatement on 16 April – reads as a case in point of the institutional phenomenon Marko Milanović describes in his recent post on EJIL:Talk!. When a political utterance would, if implemented literally, manifestly violate the law of the sea and the law of naval warfare, the legal architecture depends on the willingness and capacity of government lawyers – including, but not limited to, military legal advisers – to intervene before the order goes out to the fleet. We cannot, of course, see inside the chain of decisions that produced the substitution of “blockade of the Strait of Hormuz” with “blockade of Iranian ports and coastal areas”. But the outcome is an operation that, in the version that went to sea on 13 April, is doctrinally careful and consistent with customary international law. The geographical carve-out is the keystone. It is what reconciles the operation with Rule 99 of the San Remo Manual, what distinguishes the U.S. action from an unlawful belligerent closure of an international strait, and what preserves the parallel structure of regimes – non-suspendable innocent passage, naval blockade, and transit passage – on which the lawfulness of the whole rests. None of this resolves the larger and more serious questions raised by the conflict: the jus ad bellum legality of the American–Israeli strikes of 28 February, the April 16 expansion of belligerent visit-and-search beyond the declared geographical zone, or the legality of Iran’s own conduct in the Strait. Each of those questions deserves its own treatment, and several are receiving it. The duty of international lawyers to call out unlawful operations does not, however, relieve us of the corresponding duty to recognise lawful ones when we see them – or to identify, with some precision, the small textual choices that make the difference between the two.