Multiple species vegetative buffer strips (VBS) have been recommended as a cost-effective approach to mitigate herbicide transport in surface runoff derived from agronomic operations. However, the effect of buffer designs and species composition on reducing herbicide transport has not been well documented. An experiment consisting of three VBS designs and one cultivated fallow control replicated in triplicate was conducted to assess effectiveness in reducing herbicide transport for claypan soils. The four VBS  treatments include: (1) continuous cultivated fallow (control), (2) perennial fescue, (3) fescue with a switchgrass barrier, and (4) native vegetation (largely eastern gamagrass).  Rainfall simulation was used to create uniform antecedent soil moisture content in the plots and to generate runoff. Runoff collection equipment is installed across the plots at 1 m above (upslope) the filter strips and at 1, 4, and 8 m into the VBS treatments.  The results show a much higher proportion (40-60%) of glyphosate  transported with suspended sediments in surface runoff as compare to that of atrazine and metolachlor (<3%). All VBS significantly reduced the transport of both dissolved and sediment-bound atrazine, metolachlor and glyphosate in surface runoff.  VBS with native species were most consistently effective at reducing transport of these herbicides. Four meters of native VBS removed about 75-80% of the atrazine, metolachlor and glyphosate in surface runoff.  The additional length of native VBS had little effect on removal rates. Additionally, four meters of native species VBS resulted in much greater reductions in transport of these herbicides than eight meters of the other tall fescue VBS designs.  Thus, the implementation of native species buffers could provide desired reductions in herbicide transport with less land taken out of production.