This study investigated binding forms of cobalt (Co), copper (Cu), and lead (Pb) in 28 sediment samples from inlet to outlet of three Zambian wetlands receiving mining effluents. Use was made of a modified Tessier metal binding fractions procedure. Due to storage artefacts, the original aim of investigating the effects of redox potential (Eh) changes, starting from extremely low Eh, was suspended. Instead, use was made of the new, not often explored opportunity for replicate sample division into three categories of varying redox potential and pH. Additionally, in line with the original research aim, two sediments from each wetland were investigated for their response to increasing Eh. The results showed overall high trace metal contents, with a need for remedial actions for Co and Cu in the first, Cu in the second, and Pb in the third wetland. Rather independent of Eh and pH, Co was often found in the residual fraction (F5), as well as in the oxidizable (F4) and reducible (F3) fraction. Cu was generally dominant in F5 and F4 fractions, with low F3 prevalence, indicating a high organic matter affinity. Pb distribution among binding forms showed small variations within and across wetlands, F5, F4, and F3 fractions dominating. In the above observations, statistical analysis showed that, among the 28 sediment samples across wetlands, the influence of Eh and pH on binding forms were generally found to be not significant, being 'overruled' by other sedimentological factors. With increasing Eh, the decrease in the oxidizable (F4) fraction was smaller than expected in eight of 18 tests. The Risk Assessment Code (RAC) method, based on the exchangeable fraction (F1) plus carbonate fraction (F2), showed that some sediments turned from "unsafe" to "safe," and vice versa, with increasing Eh. The "total metals method" does not show bioavailability, whereas RAC does not use the metal contents. Thus, the two methods should be used together to improve the prediction of potential toxicity.