One of today’s greatest conservation challenges is the proliferation of invasive species. A particularly ecologically and economically damaging invasive species is the zebra mussel (Dreissena polymorpha). First arriving in North America in the Great Lakes region, zebra mussels have expanded across most of the eastern half of the US and pose an ongoing concern for potential introduction and establishment in the western United States. Zebra mussels are already present within parts of Texas and if current trends persist, there are concerns that this invasion front may reach New Mexico. Our study applies network analysis and habitat suitability indices to identify the lakes most critical to the spread and establishment of zebra mussels in Texas and New Mexico’s waterbodies. Our water quality-based habitat suitability indices identified many lakes in east Texas with no or low suitability for zebra mussel infestation. Nearly all lakes in other regions of the study area were classified as having moderate to high suitability. We created a network consisting of lake nodes and connecting roadway edges, using a maximum edge length of 363 km based on Texas boater movement surveys. Using graph analysis, we identified lakes critical to the spread of zebra mussels by acting as hubs (lakes connected to a high number of nearby lakes), steeping stones (lakes that facilitate a disproportionate amount of dispersal within the network), or cutpoints (lakes when removed from the network, increase the network fragmentation). Hubs were concentrated in eastern and central Texas, with connectivity rapidly declining further west as lakes became more isolated. However, due to high connectivity throughout the overall network, no cut points were identified. Stepping stones were distributed throughout the study area, with no obvious pattern. We then compiled a list of lakes that were found at least twice among the top 20% of most suitable habitats, hubs, and stepping stones. These 16 lakes represent the sites most critical to zebra mussel spread and establishment. By identifying these critical lakes, we can help resource managers guide the allocation of limited time and resources for early detection and outreach campaigns. Our study also aimed to assess the overall infestation risk of Texas and New Mexico lakes to facilitate early detection and mitigation strategies. We assessed this risk by combining a habitat suitability index (HSI) model and a production-constrained gravity model. This approach allows us to synthesize the risk of introduction (gravity) and the risk of establishment (HSI) for an overall risk of infestation. We examined the impact of habitat suitability, zebra mussel survival, distance coefficients, maximum boater travel distances, and infested boat thresholds on model performance. We modeled the zebra mussel expansion from the initial invasion of Texas in 2009 to the year 2040, using the existing historical record to evaluate our models. The best performing models were those model runs that successfully predicted at least 70% of real-world lake infestations, while incorrectly predicting as infested no more than 30% of real-world noninfested lakes. Our best performing models identified many lakes in central Texas as at risk of infestation but predicted modest infestations in west Texas and New Mexico. Within these models, new infestations greatly decreased over time, with the majority of models predicting no new infestations after 2022. These model predictions can guide resource managers when allocating limited time and resources for early detection, mitigation, and outreach campaigns.