Keeping trains moving is no longer enough
Ukraine’s railways are still moving passengers and freight despite repeated Russian attacks. That operational success now risks obscuring a deeper problem. Ukrzaliznytsia reports that Russia has hit 304 locomotives and 120 passenger wagons across the country this year. Separately, strikes near Poland show that Russia’s campaign has reached important border nodes. A railway can keep running while repeated attacks damage more rolling stock and expose border links.
“A railway can keep running while repeated attacks damage more rolling stock and expose border links.”
The pressure intensified in mid-September. A Russian drone struck railway infrastructure near Dorohusk–Yahodyn on 13 September. Ukrainian Railways said a diplomatic train might have been the intended target, but that assessment has not been independently established. On 16 September, drones attacked a passenger train in Mykolaiv region. Officials said all 168 passengers and crew were evacuated without injury. The acting regional governor said the train was attacked twice more after the first strike. A locomotive was also hit at Kovel, a rail point near Poland.
These incidents differ from an isolated strike that crews can repair and route around. Damage to a locomotive removes traction capacity until it returns to service or is replaced. Repeated attacks on border approaches can also threaten passenger links and commercial traffic at crossings. The effect can accumulate even when an individual incident produces no casualties and only a short visible delay.
Key takeaways
- Ukrzaliznytsia says Russia has hit 304 locomotives and 1,700 railway facilities in 2026, although the totals are not independently audited and do not reveal repair times or reserve capacity.
- Safe evacuation and rapid service restoration remain important, but they can coexist with cumulative rolling-stock damage and attacks reaching crossings near Poland.
- Ukraine’s 15 September contracting plan should be assessed after deployment by comparing locomotive damage per attack and border-node downtime at protected and similarly exposed unprotected sites.
The operator’s figures point to cumulative damage
Ukrzaliznytsia told Reuters that Russia attacked 842 railway facilities in 2024 and 1,206 in 2025. Its partial-year figure for 2026 is already 1,700. The operator also reported 120 passenger wagons, 304 locomotives and 43 stations hit this year. Foreign Minister Andrii Sybiha separately said more than 500 locomotives had been damaged or destroyed since Russia launched its full-scale invasion, with more than half of that damage occurring in 2026.
The two locomotive figures should not be merged. They may count different types of damage, repeat strikes or reporting periods, and neither has been independently audited. The cited reports do not provide the geographic distribution of the hits, the number of locomotives returned to service or the size of the available reserve. Even with those limitations, the reported increase is too large to assess railway performance only by whether trains resume after an attack.
Russia is the aggressor, and it has attacked civilian rail assets that Moscow claimed were also used for military cargo. That Russian claim about western rail infrastructure does not remove responsibility for attacks on a network used extensively by civilians, especially after Russia’s invasion closed Ukrainian airspace. Nor does it establish that the diplomatic train was the specific target. Ukraine’s effort to protect the railway is defensive. It is intended to preserve civilian mobility and the national logistics on which the country depends under invasion.
Protection contracts need measurable outcomes
On 15 September, President Volodymyr Zelenskyy said there would be additional contracts for systems to strengthen protection of railways, border crossings and other Ukrzaliznytsia facilities, and that decisions on those contracts were already in place. He also said he had discussed possible state-bank support with the prime minister and finance minister. The announcement identifies the right operational problem, but it gives no system specifications, suppliers, budget, contracting timetable, delivery dates or list of protected sites.
The contracting plan is an input, not evidence that vulnerability has fallen. A useful assessment would compare changes in locomotive damage per attack and border-node downtime at protected sites with similarly exposed unprotected sites. Where the data allow, the comparison should account for weapon type, attack intensity and target type. Repair turnaround and spare capacity would help assess whether damaged locomotives are returning to service fast enough. Cancelled services, delay hours and freight throughput would measure effects on users. Evacuation outcomes still matter, but a perfect evacuation record cannot establish the condition of the fleet and routes behind it.
Public reporting need not reveal the location or capability of defensive systems. Ukrzaliznytsia could publish monthly national totals for rolling stock hit, repair turnaround, cancellations and major delay hours. It could also release anonymised attack-event counts and outcomes for protected and comparison groups. Aggregated figures for border-crossing downtime and freight flows would support evaluation without giving Russia a site-level map of protection.
The available evidence supports a shift from recovery after separate incidents to protection against cumulative damage to rolling stock and border nodes. Ukraine has kept trains moving and evacuated passengers safely, but the reported locomotive hits and 2026 facility count show that continuity alone is no longer an adequate measure. The judgment would change if audited data showed stable reserve capacity and lower locomotive damage per attack and border-node downtime at protected sites than at comparable unprotected sites, after accounting where possible for differences in attacks. It would also change if the operator totals proved to include extensive duplication. Until such evidence is available, the 15 September contracting plan remains a response to the problem, not proof that protection has improved.
Share this analysis
Mara Ionescu
Research ContributorInformation resilience, foreign influence narratives, and regional security in Romania, Ukraine, and the Republic of Moldova.
View contributor profile →
