Understanding how non-native plants establish and spread in new locations is increasingly important in the face of the homogenization of the global flora. To investigate both successful and unsuccessful plant introductions, we take a historical perspective, using digitized herbarium specimens to document the plants inadvertently introduced into Northeastern North America from Europe, Asia, Africa, South and Central America as a result of historical shipping trade. Specifically, we investigate the plants introduced into New Jersey through the deposition of solid ships’ ballast during the 19th century. Such ships’ ballasts were used to balance weight on sailing ships depending on the transported cargo, and included materials such as rocks, sand, soil fill, and used building materials. When steamships started to replace sailing ships for oceanic trade, solid ballast was replaced with water ballast (the type of ballast that is used today). We used digitized herbarium specimens from 75 herbaria to study the establishment and spread of 264 vascular plants associated with ballast deposition in New Jersey (a total of 6433 herbarium specimens from 1800-2022). Using temporal and spatial analyses of species retention and geographic spread we defined the trajectories these plants followed and quantified disappearance rate (species lost through time), survival (recollection in New Jersey), and geographic dispersion through time from 1800s to today (geolocation of observations and specimens). Four distinct trajectory groups based on species survival and spread were identified: Waifs (species only present during active ballast import times; 32% of total ballast species), Short-Term Introductions (species disappeared quickly after active import ended; 20%), Established & Limited Spread (species still survives locally, 30%), and Established & Widespread (species now geographically widespread, 18%). Species disappearance rate was high during the period of active ballast deposition and decreased following the end of ballast deposition (around 1900). We also demonstrate that the disappearance rate and spatial analysis are surprisingly robust to collection bias in the herbarium record. This is one of very few studies investigating the successes and failures of a large set of inadvertently introduced species through time and highlights the utility of herbarium specimens in studying plant introductions on large time scales and on a global scale.