To improve prediction of the heavy metals between deficiency and toxicity in soil systems especially alluvial ones, it is critical to understand time dependent sorption and desorption behavior. In this paper, desorption behaviours of cobalt in three cultivated alluvial soils (Typic torrerts) mainly varying in their clay contents and other soil properties were studied using Electrical Stirred Flow Unit (ESFU). The preliminary experiment was conducted to understand Co retention in alluvial soils. Soil samples were shaken in ESFU with three concentrations namely; 0, 25, 75 mg kg-1 representing deficient, nutrient and toxic levels of Co until reaching equilibrium, and the Distribution Coefficient (DC) was calculated. In desorption experiments, the same equilibrated soil samples were divided into three parts and incubated for two days, two weeks and eight weeks. After each incubation time, the kinetics of Co desorption for the studied soil samples were conducted using the ESFU for the same shaking periods i.e. 1 min -24 hrs. The pure solutions collected from the unit after each reaction time for both adsorption and desorption were analyzed for their Co concentrations and the kinetic data obtained were fit to two empirical and theoretical rate models -- i.e. Elovich and first-order equations -- and the kinetic parameters of these models were calculated and analysed to represent cobalt status in the alluvial soils. The obtained results indicated that (DC), derived from created model used in this study, was influenced by soil characteristics especially pH and clay content in these soils. Also, data emphasized movement of Co toward slower adsorption reactions with increasing period of soil incubation. However, some cautions should be considered in interpreting such parameter. Results indicate that Co desorption is rate limiting as evidenced by initial fast reaction followed by a decrease in the outflow concentration. Period of incubation time and rate of Co applied in addition to soil characteristics influenced fitting data to given kinetic models. However, both equations used well described Co desorption data with priority to empirical one. The kinetic parameters were increased by increasing rate of Co applied and decreased with increasing incubation period. In Co treated soils, kinetic parameters were mainly influenced by soil clay content since k parameter of 1st order kinetic equation were decreased in the order of increasing clay contents of the studied soils after 6 hrs of ESFU working. Apparently, the higher specific surface and the slightly higher pH of the alluvial soil contributed in capacity and rate of Co-release. Different mechanisms influencing Co desorption from alluvial soils were discussed.