The effects of soil manganese on soil cobalt availability for plant uptake were studied systematically through various chemical extractants, sequential chemical fractionation, glasshouse pot trials and field monitoring programs. Results have led to a better understanding of the nature of the relationship between cobalt and manganese in soil and its influence on cobalt uptake by plant. As assessed by different chemical extractants, cobalt has strong correlations with manganese in the New Zealand soils studied. A significant proportion of soil cobalt is associated with soil manganese and shows similarities in its relationships with soil pH, soil moisture conditions, soil iron and aluminium. The total amount of cobalt in soils is tightly related to soil parent materials. The downward movement of cobalt is not significant for New Zealand soils. Over a wide range of New Zealand soil types, EDTA extraction appear to be too strong to assess cobalt availability for plant uptake because it dissolves cobalt, simultaneously with manganese, from soil solid forms which may not be available for plant uptake, and may obscures the pH effect on the availability of soil cobalt. As assessed by CaCl???extraction, only small proportions of cobalt and manganese are present in exchangeable forms in the New Zealand soils studied. Normal air-drying of soils increases the exchangeable cobalt and manganese significantly. The concentrations of soil exchangeable cobalt and manganese show similarities with respect to changes in soil pH. A sequential chemical fractionation has been applied successfully to assess soil cobalt and manganese status, and their relationships in individual fractions in a wide range of New Zealand soils. Compared with the soil oxide and residual fractions there are relatively small proportions of soil cobalt in the organic-bound fraction. Compared with other soil groups, soils derived from pumice parent materials hold relatively high proportions of total cobalt in organic-bound forms. There are considerable proportions of manganese existing in free manganese oxide forms. The concentrations and proportions of cobalt are substantially lower in the free manganese oxide fraction compared with those of cobalt in the two iron oxide fractions. Substantial proportions of cobalt are present in forms associated with soil amorphous and crystalline iron oxides. The largest proportion of soil cobalt is found in the residual fraction. Based on the studied New Zealand soils with large variations of soil manganese status and other soil properties, and under glasshouse conditions, it is demonstrated that soil manganese status has a close relationship with soil cobalt availability. Soils with high manganese contents have a high probability of strong fixation of soil cobalt and show negligible responses to cobalt fertiliser treatment. However, differences in the response of ryegrass pasture cobalt concentrations to the amendment of soil manganese status indicate that, apart from soil manganese status, other soil properties, such as the status of soil secondary iron- and aluminium-oxide minerals, may also influence the retention of soil cobalt for plant uptake. Soil moisture and soil pH also present appreciable effects on plant cobalt concentration although the results are extremely variable. Compared to glasshouse trials, much more variable results were obtained for pasture cobalt concentrations in field monitoring programs. The high sensitivity of soil cobalt availability to changes in soil pH and soil moisture conditions is a common characteristic that may be the main cause of the substantial changes in pasture cobalt over short periods. Under field conditions, soils with lower retention capacities for cobalt are likely to show both high average pasture cobalt concentrations and high seasonal variations. Generally under field conditions, there were poor correlations between pasture cobalt and the contents of cobalt and/or manganese in various soil chemical extractions. In the present study, it was demonstrated that older and/or well developed soils with high contents of manganese, iron and aluminium minerals in their secondary minerals generally showed strong fixation for soil cobalt, native and/or applied. In such soils, the cobalt availability would be controlled to a large extent by the soil manganese status. However, younger soils or soils with low manganese, iron and aluminium in their secondary minerals showed weak fixation for soil cobalt. In such soils, cobalt availability would be dominated by the status of native cobalt, and the applications of cobalt fertiliser are more likely to obtain an appreciable response.