Abstract. We assess the potential damage and smoke production associated with the detonation of small nuclear weapons in modern megacities. While the number of nuclear warheads in the world has fallen by about a factor of three since its peak in 1986, the number of nuclear weapons states is increasing and the potential exists for numerous regional nuclear arms races. Eight countries are known to have nuclear weapons, 2 are constructing them, and an additional 32 nations already have the fissile material needed to build substantial arsenals of low-yield (Hiroshima-sized) explosives. Population and economic activity worldwide are congregated to an increasing extent in megacities, which might be targeted in a nuclear conflict. We find that low yield weapons, which new nuclear powers are likely to construct, can produce 100 times as many fatalities and 100 times as much smoke from fires per kt yield as previously estimated in analyses for full scale nuclear wars using high-yield weapons, if the small weapons are targeted at city centers. A single "small" nuclear detonation in an urban center could lead to more fatalities, in some cases by orders of magnitude, than have occurred in the major historical conflicts of many countries. We analyze the likely outcome of a regional nuclear exchange involving 100 15-kt explosions (less than 0.1% of the explosive yield of the current global nuclear arsenal). We find that such an exchange could produce direct fatalities comparable to all of those worldwide in World War II, or to those once estimated for a "counterforce" nuclear war between the superpowers. Megacities exposed to atmospheric fallout of long-lived radionuclides would likely be abandoned indefinitely, with severe national and international implications. Our analysis shows that smoke from urban firestorms in a regional war would rise into the upper troposphere due to pyro-convection. Robock et al. (2007) show that the smoke would subsequently rise deep into the stratosphere due to atmospheric heating, and then might induce significant climatic anomalies on global scales. We also anticipate substantial perturbations of global ozone. While there are many uncertainties in the predictions we make here, the principal unknowns are the type and scale of conflict that might occur. The scope and severity of the hazards identified pose a significant threat to the global community. They deserve careful analysis by governments worldwide advised by a broad section of the world scientific community, as well as widespread public debate.