In Florida, waterhyacinth (Pontederia [Eichhornia] crassipes) (Martius) Solms Laubach (Pontederiaceae) is primarily controlled by herbicides, but overall control is enhanced by insect biological control agents that decrease growth and reproduction and slow regrowth. However, herbicide applications often disrupt the biological control agent populations when applied indiscriminately. Previous studies identified the utility of preserving populations of biological control agents in the vicinity of herbicide treated areas by establishing refuges for the insects. The objectives of this experiment were to quantify the impact of insect refuges, using groups of untreated P. crassipes within treated mats, on the regrowth of the new mat and the ability of biological control agents to persist following an herbicide treatment. Pontederia crassipes mats were grown with and without biological control agents, then treated with 2 concentrations of the herbicide penoxsulam. Plant growth metrics and biological control agent densities were monitored pre- and post-treatment and compared using ANCOVAs and non-parametric Kruskal-Wallis tests. Although the systemic activity of penoxsulam in the water column prevented the establishment of refuges in this study, biological control agent populations persisted following herbicide applications without additional releases and were able to remain at field densities after the decay and loss of P. crassipes. The treatment with no biological control agents and only half-label rate herbicide grew more densely, had greater dry weight biomass, higher relative growth rate, and produced more inflorescences than the plants in treatments with biological control agents. The half-label herbicide and biological control agent treatment performed as well as both treatments with label rate herbicide, and with and without biological control agents, respectively, in lowering P. crassipes density, final dry weight biomass, and relative growth rate. Although the concept of refuge systems at operational field scales requires further study, demonstrating the ability of biological control agents to persist even on sprayed mats is a necessary first step in determining the temporal and spatial factors that might influence the utility of such refuges.