Sowing native seeds is a common approach to reintroduce native plants to degraded systems. However, this method is often overlooked in wetland restoration despite the immense global loss of diverse native wetland vegetation. Developing guiding principles for seed-based wetland restoration is critical to maximize native plant recovery, particularly in previously invaded wetlands. Doing so requires a comprehensive understanding of how restoration manipulations, and their interactions, influence wetland plant community assembly. With a focus on the invader Phragmites australis, we established a series of mesocosm experiments to assess how native sowing density, invader propagule pressure, abiotic filters (water and nutrients), and native sowing timing (i.e., priority effects) interact to influence plant community cover and biomass in wetland habitats. Increasing the density of native seeds yielded higher native cover and biomass, but P. australis suppression with increasing sowing densities was minimal. Rather, community outcomes were largely driven by invader propagule pressure—Phragmites australis densities of ≤ 500 seeds/m² maintained high native cover and biomass. Low-water conditions increased the susceptibility of P. australis to native competition. Early sowing of native seeds showed a large and significant benefit to native cover and biomass, regardless of native sowing density, suggesting that priority effects can be an effective restoration manipulation to enhance native plant establishment. Given the urgent wetland restoration need combined with the limited studies on seed-based wetland restoration, these findings provide guidance on restoration manipulations that are grounded in ecological theory to improve seed-based wetland restoration outcomes.