8 pages, 4 figures, 4 tables, 32 references.Chloridazon and lenacil are herbicides widely used in sugar beet crops of the Guadalquivir river marshes, southwest Spain. However, little is known on the behaviour and environmental impact of these herbicides in these soils of very high clay and salt contents. Laboratory and field experiments were performed to evaluate the role of different soil processes on the fate of both herbicides. Chloridazon sorbed to a higher extent than lenacil and sorption was less reversible. Laboratory experiments suggested that the singular altered-illite present in the soil was mainly responsible for the higher sorption of chloridazon. Dissipation experiments at 20 8C showed faster dissipation of chloridazon than lenacil, although chloridazon had a longer lag phase than lenacil, followed by a very rapid dissipation. No herbicide was detected in leachates from undisturbed soil columns, and the mobility within the soil column was very low. The lack of preferential water flow in the clayey soil, confirmed by bromide movement, was partly responsible for this. This favoured sorption and degradation processes. Half-lives were also shorter than in the static dissipation experiments. Field measurements also showed very low mobility of both herbicides in the soil profile and similar t1/2 values than those calculated from soil columns in the laboratory. When compared to chloridazon, higher amounts of lenacil were detected and at greater depths, likely due to the higher persistence, lower sorption and higher sorption reversibility of lenacil. No traces, neither of lenacil nor chloridazon were detected in water from drainage pipes in the field, indicating a low risk of groundwater contamination by these herbicides under the studied field conditions.This work was financed by the FP5 EU Project APECOP, ref. QLK4-CT-1999-01238, and by Junta de Andalucıa (PAI research group RNM 0124).Peer reviewed