A search for clay-surfactant based formulations with high percentage of the active ingredient, which can yield slow release of active molecules is described. The active ingredients were the herbicides metribuzin (MZ), mesotrione (MS) and flurtamone (FL), whose solubilities were examined in the presence of four commercial surfactants; (i) neutral: two berols (B048, B266) and an alkylpolyglucoside (AG6202); (ii) cationic: an ethoxylated amine (ET/15). Significant percent of active ingredient (a.i.) in the clay/surfactant/herbicide formulations could be achieved only when most of the surfactant was added as micelles. MZ and FL were well solubilized by berols, whereas MS by ET/15. Sorption of surfactants on the clay mineral sepiolite occurred mostly by sorption of micelles, and the loadings exceeded the CEC. Higher loadings were determined for B266 and ET/15. The sorption of surfactants was modeled by using the Langmuir-Scatchard equation which permitted the determination of binding coefficients that could be used for further predictions of the sorbed amounts of surfactants under a wide range of clay/surfactant ratios. A possibility was tested of designing clay-surfactant based formulations of certain herbicides by assuming the same ratio between herbicides and surfactants in the formulations as for herbicides incorporated in micelles in solution. Calculations indicated that satisfactory FL formulations could not be synthesized. The experimental fractions of herbicides in the formulations were in agreement with the predicted ones for MS and MZ. The validity of this approach was confirmed in in vitro release tests that showed a slowing down of the release of a.i. from the designed formulations relative to the technical products. Soil dissipation studies with MS formulations also showed improved bioactivity of the clay-surfactant formulation relative to the commercial one. This methodological approach can be extended to other clay-surfactant systems for encapsulation and slow release of target molecules of interest.