Propagule pressure is fundamental to invasion success, yet our understanding of its role in the marine domain is limited. Few studies have manipulated or controlled for propagule supply in the field, and consequently there is little empirical data to test for non-linearities or interactions with other processes. Supply of non-indigenous propagules is most likely to be elevated in urban estuaries, where vessels congregate and bring exotic species on fouled hulls and in ballast water. These same environments are also typically subject to elevated levels of disturbance from human activities, creating the potential for propagule pressure and disturbance to interact. By applying a controlled dose of free-swimming larvae to replicate assemblages, we were able to quantify a dose–response relationship at much finer spatial and temporal scales than previously achieved in the marine environment. We experimentally crossed controlled levels of propagule pressure and disturbance in the field, and found that both were required for invasion to occur. Only recruits that had settled onto bare space survived beyond three months, precluding invader persistence in undisturbed communities. In disturbed communities initial survival on bare space appeared stochastic, such that a critical density was required before the probability of at least one colony surviving reached a sufficient level. Those that persisted showed 75% survival over the following three months, signifying a threshold past which invaders were resilient to chance mortality. Urban estuaries subject to anthropogenic disturbance are common throughout the world, and similar interactions may be integral to invasion dynamics in these ecosystems.