Biopolymers, such as xanthan gum, have been shown to reduce erosion of earthen embankments. In non-cohesive soils, biopolymers absorb water forming a viscous matrix between grains. When clay particles are present, biopolymers also interact with clays further increasing the resistance to erosion. It is unclear, however, how biopolymer treated soil behaves during embankment overtopping and how it impacts breach evolution. Further, questions remain on how erodibility and permeability of biopolymer treated soil varies with soil gradation. Laboratory experiments were performed at the University of South Carolina to study breach evolution in biopolymer treated, compacted dams built with a mixture of 0.35 mm sand and silt. In the experiments, dam height was 20 cm and dam width 27 cm, which spanned the entire flume width. The upstream reservoir level was maintained at the dam crest until failure. Concentrations of silt and xanthan gum were systematically varied, and dam failure times were compared. Experiments showed that failure time of sand dams sharply increased as biopolymer concentration increased from 0% to 0.2% by soil mass. Increasing silt concentration in sand dams from 0% to 25% only mildly increased dam failure time, illustrating that biopolymers were much more effective than silt at reducing erosion. The experiments further showed that when biopolymers were added to the sand-silt mixture, dam failure was noticeably delayed compared to the case of biopolymer treated sand dams. These results suggest that biopolymers can be a useful, low-cost additive to reduce erodibility of earthen embankments, but further research is necessary to assess the applicability at the field scale.