The Cold War was fought by every method except the one the rivals feared most. After the Cuban Missile Crisis of October 1962 brought Washington and Moscow within hours of thermonuclear war, both superpowers accepted an unwritten rule: no direct fighting between American and Soviet forces. They kept competing anyway — for territory, for allies, for prestige — by arming, training, and sometimes directing local armies. The result was a chain of conflicts from the Sahara to the Mekong in which local grievances, regional rivalries, and superpower interests fused into wars that killed millions and outlasted the Cold War itself.
The most consequential of these proxy wars are treated in depth in the four in-depth articles of this overview: the Angolan Civil War, the Ogaden War, the Nicaraguan conflict between the Sandinistas and the Contras, and the Cambodian-Vietnamese War.
How Proxy Wars Worked
The phrase “proxy war” obscures more than it reveals. Most Cold War interventions were not clean cases of superpower X fighting superpower Y through local client Z. They were layered, ad hoc, and frequently improvised under conditions of uncertainty. The CIA did not direct UNITA; Jonas Savimbi ran his own war. The Kremlin did not order Castro to send 50,000 troops to Angola; Castro decided and then asked Moscow to underwrite the bill. The Reagan White House did not want war with Nicaragua’s Sandinistas so much as it wanted them gone, and funded proxies whose command it could not fully control.
What the superpowers did provide was decisive: weapons, training, intelligence, logistics, satellite imagery, hard currency, diplomatic cover, and — in extremis — combat troops (Cuban in Angola and Ethiopia, Vietnamese in Cambodia, Soviet “advisers” almost everywhere). Local patrons traded political independence for material survival. The arrangements were unequal and often exploitative, but the local actors were rarely puppets. They had their own programs, factions, and calculations.
Odd Arne Westad’s The Global Cold War (2005) is the standard treatment of this system. Westad argues that the Third World wars were not side-shows but the main event of the post-1945 Cold War — the place where the contest between capitalism and communism was actually adjudicated, and where the superpowers’ rhetoric about modernization, sovereignty, and revolution was tested against realities of village politics, ethnic violence, and failed harvests. Earlier accounts, including John Lewis Gaddis’s Strategies of Containment (1982), emphasized state-level strategy; the revisionist literature of the 1990s and 2000s, exemplified by Westad, Piero Gleijeses, and others, pushed the lens downward to the local and the personal.
The Cuban–Soviet–American Triangle
The proxy war system had a strange geometry. The United States and the Soviet Union were nominally the principals, but the actor that did most of the on-the-ground fighting of the late 1970s was Cuba. Between 1975 and 1989, Fidel Castro deployed Cuban troops to Angola (peaking near 50,000), Ethiopia (around 17,000 in 1978), and a number of smaller commitments. Havana was not Moscow’s tool: Castro repeatedly committed Cuban forces without prior Soviet approval and extracted Soviet subsidies afterward. The arrangement resembled a marriage of convenience more than a chain of command.
Moscow benefited enormously. Cuban troops relieved the Kremlin of the political risk of deploying Soviet conscripts abroad, and they fought well. The cost, ultimately, was structural: when the USSR collapsed in 1991, Cuba’s external empire evaporated within months. The same dependency worked in reverse for the United States, which armed and financed clients (the Contras, UNITA, the mujahideen, Mobutu) without ever commanding them, and which often discovered that client interests diverged sharply from Washington’s.
Africa as a Cold War Battleground
Africa was the principal theater of Soviet–American–Cuban confrontation in the 1970s. The Portuguese collapse of April 1974 opened Angola and Mozambique to immediate proxy competition. In Angola, the Soviet Union and Cuba backed the MPLA, while the United States, apartheid South Africa, and Zaire’s Mobutu backed the FNLA and UNITA. The decisive Cuban intervention from late 1975 (Operation Carlota) routed the South African armored thrust at the Battle of Bridge 14 in March 1976 and secured Luanda. Gleijeses’s Conflicting Missions (2002) and Visions of Freedom (2013) are the indispensable accounts, drawn from Cuban, American, and South African archives.
The Horn of Africa followed two years later. Somalia’s Siad Barre, financed and armed by Moscow through 1977, invaded the Ogaden in July 1977. When Barre refused to break with Washington and Saudi Arabia, the Soviets pivoted: in November 1977, Brezhnev authorized a massive airlift to Ethiopia. Some 17,000 Cuban troops and Soviet advisers turned the war around by March 1978. The reversal was one of the most abrupt in Cold War history and remains the clearest case of a superpower client being abandoned mid-war.
Latin America as a Cold War Battleground
Latin America was a different kind of theater — geographically American, ideologically contested, and shaped by the memory of the 1959 Cuban Revolution and the 1954 Guatemalan coup. Washington’s preferred instrument was covert action. The CIA helped overthrow Salvador Allende in Chile (September 1973), supported the 1981 Contra war in Nicaragua, and tolerated (and sometimes collaborated with) Operation Condor, the 1975–1983 transnational repression scheme run by the Southern Cone dictatorships. Soviet and Cuban involvement was real but small — the Sandinistas received arms, oil, and advisers, but the USSR was careful not to repeat its Afghanistan overstretch in a region it could not easily reach.
The most significant Central American conflict of the Cold War — the Nicaraguan civil war of 1979–1990 — is examined in detail in the Sandinistas and the article Contras. It is also the conflict in which American domestic politics most directly shaped proxy-war policy: the Boland Amendments (1982, 1984) prohibited U.S. funding of the Contras, and the Reagan administration’s determination to fund them anyway produced the Iran-Contra scandal of November 1986.
Asia as a Cold War Battleground
Southeast Asia saw the longest and bloodiest of the proxy wars. Vietnam, itself a Cold War client of Moscow after 1975, invaded Cambodia on December 25, 1978 to overthrow the Khmer Rouge, whose border attacks and domestic genocide made it, in Hanoi’s calculation, a Chinese-backed threat. The fall of Phnom Penh on January 7, 1979 inaugurated a Vietnamese occupation that lasted until September 1989. The war produced the longest sustained refugee crisis of the Cold War, the eventual Paris Peace Accords of October 23, 1991, and a slow Cambodian reconstruction that began with UN-supervised elections in May 1993.
The Vietnam War itself (1955–1975) and the Sino-Vietnamese border war of February 1979 — Beijing’s punitive invasion launched in the same month as Vietnam’s Cambodian operation — are the larger context. The Cambodian-Vietnamese War article treats the proxy dimension in detail, drawing on the work of Ben Kiernan on the Khmer Rouge genocide.
Human Cost
The arithmetic of Cold War proxy wars is grim. The Cambodian genocide killed an estimated 1.5 to 2 million people, roughly a quarter of the country’s population. The Angolan Civil War, counted through its end in 2002, killed perhaps 500,000 and displaced four million. The Nicaraguan war of the 1980s produced around 30,000 Nicaraguan dead and an economy wrecked by inflation and U.S. trade embargo. The Ogaden War killed tens of thousands and helped set Somalia on the path to state collapse in 1991. Add the Korean War, Vietnam, the Soviet-Afghan War, the Ethiopian-Eritrean conflict, Mozambique, and the smaller commitments — the cumulative death toll from proxy-war violence runs into the low millions, with three or four times that number of refugees, internally displaced, and wounded.
The political costs compounded the demographic ones. Several proxy-war states (Angola, Ethiopia, Cambodia, Nicaragua, Afghanistan) emerged from the Cold War with collapsed economies, broken institutions, and a generation of fighters who had no trade but violence. The 1990s civil wars in Somalia, Liberia, Sierra Leone, and the post-2001 instability across the Sahel are partly downstream of Cold War-era arming patterns. Latin American states that had hosted U.S.-backed security forces in the 1970s spent the 1990s digging up mass graves. Cambodia’s Extraordinary Chambers in the Courts of Cambodia convicted senior Khmer Rouge leaders of genocide as late as 2018.
Long-Term Consequences
The proxy wars of the Cold War also produced institutional legacies that outlived the conflict. The U.S. practice of large-scale covert action — armed intervention, regime support, and information warfare run outside normal diplomatic channels — was institutionalized in the 1980s and reappeared, with modifications, in Iraq, Libya, and Syria in the 2000s and 2010s. The Soviet Union’s overcommitment to distant clients in 1979–1989 helped accelerate the fiscal crisis that broke it in 1991. Cuba emerged from the Cold War with a hardened, professionalized military and a foreign-policy doctrine, codified in the 1986 Concepto de la Guerra de Todo el Pueblo tradition, of “internationalist” combat that survives in Russian and Venezuelan security cooperation in the 2020s.
For scholars and practitioners, the basic lesson of the proxy wars is that the “other side” in such a conflict is rarely a single adversary. Local factions defect, change patrons, and pursue their own wars within wars. The post-Cold-War period did not end proxy conflict so much as reformat it: the same patron–client architecture has resurfaced in Syria, Yemen, and Ukraine in the 2010s and 2020s, sometimes with the same outside powers and often with the same logic. Westad’s central insight — that the Third World wars of 1945–1991 were constitutive, not derivative, of the Cold War — has aged well. So has Gleijeses’s caution that the patron–client relationship runs in both directions, and that the clients often have minds of their own.