Fourteen soils comprising A and B horizons, from Canterbury, New Zealand were examined for the sorption/desorption behaviour of added and native soil Zn. These soils were selected to represent a wide range of soil properties likely to be of major importance in controlling the sorption or desorption process of Zn from soils. The DTPA-extractable Zn (presumably available Zn) of these soils ranged from 0.12 to 15.98 ??g g????? soil. The amounts and patterns of added and native soil Zn desorbed varied between the different soils studied. Greater concentrations of native Zn were desorbed after five consecutive desorption periods from the surface soils (from 0.33 /??g g????? to 2.09 /??g g????? soil) than from the sub-surface soils (from 0.11 /??g g-l to 0.57 ??g g????? soil). Sorption of Zn was consistently lower in sub soil samples compared with surface soils. In contrast to native Zn desorption, the added Zn desorbed (%) was higher in sub surface soils than in surface soils. Cation exchange capacity and organic C were found to be the dominant soil variables contributing towards sorption or desorption of Zn. Cation exchange capacity itself accounted for most of the variation (48 to 62 %) in native Zn desorption, Zn sorption, and added Zn desorption. However, inclusion of clay and Mn oxides in the case of native Zn desorption; pH and Mn oxides in the case of Zn sorption; and clay, soil pH and amorphous Al oxides in the case of added Z desorption explained nearly 85 to 97 % of the variation between these Canterbury soils. Desorption of Zn was found to be reversible in soils having coarse texture and it closely followed the original sorption isotherm suggesting that in these soils, desorption reactions could be described by the sorption isotherm. However, for a soil with comparatively high clay content and a high CEC, desorption of Zn was only partially reversible and there was a marked hysteresis effect between sorbed and desorbed Zn. In the majority of soils studied, the longer the period of contact time of Zn with soil, the smaller was the Zn desorption (20 to 36.5 % reduction after 90 days). Zinc sorption/desorption varied widely depending on soil pH. Sorption of Zn increased with increase in pH and at pH near 6.5 most of the added Zn was sorbed by all the soils studied. Desorption of native and added Zn decreased with increase in pH and became very low as pH approached near neutral. The decrease, in both native Zn desorption as well as added Zn desorption, was larger in the pH range of 4.3 to 5.4 than in the range 5.4 to 6.5 in all the four soils studied. An examination of sorption/desorption isotherms indicated that the extent of reversibility decreased as the pH of soils increased. At pH near 4.3, Zn desorption closely followed the original sorption isotherm irrespective of the concentration of Zn added to the soils studied. As the pH of these soils was raised to 6.4 or above, there was a marked hysteresis effect between sorbed and desorbed Zn. The desorption of added Zn (5??g g????? soil) and native Zn from the soils over a range of pH values was not affected by the addition of fertilizer P (up to 50 kg ha????? ) suggesting that there is a low probability of P-Zn interaction in the soil system itself. Resin membranes appear to have considerable advantages for studying the kinetics of Zn desorption compared with chemical extractants such as 0.01 M Ca(NO???)??? or 0.005 MDTPA. The rates of native and added Zn desorption using both types Ion exchange and chelating of membranes at lower pH were rapid initially and gradually declined with time. The kinetics of native Zn desorption were best described by simultaneous first-order and pseudo first order models, and added Zn desorption by first-order and pseudo first order models. The parabolic diffusion model also gave reasonably good fits for both native and added Zn desorption. As the desorption data could be predicted by several different types of kinetic models, it is probable that in the heterogeneous soil system more than one type of mechanisms are likely to be involved. With increasing pH from 4.3 to 6.5, the Zn desorption rates in the soils were decreased. Increasing the length of contact period also substantially decreased the rates of Zn desorption. Irrespective of the length of contact time of Zn with soil, simultaneous first order, pseudo first order and parabolic diffusion gave good fits for the rate of Zn desorption.