The establishment of phytoextraction crops on highly contaminated soils can be limited by metal toxicity. A recent proposal has suggested establishing support crops during the critical initial phase by metal immobilization through soil amendments followed by subsequent mobilization using elemental sulphur to enhance phytoextraction efficiency. This 'combined phytoremediation' approach is tested for the first time in a pot experiment with a highly contaminated soil. During a 14-week period, relatively metal-tolerant maize was grown in a greenhouse under immobilization (before sulphur (S) application) and mobilization (after S application) conditions with soil containing Cd, Pb and Zn contaminants. Apart from the control (C) sample, the soil was amended with activated carbon (AC), lignite (Lig) or vermicompost (VC) all in two different doses (dose 1~45 g additive kg-1 soil and dose 2~90 g additive kg-1 soil). Elemental S was added as a mobilization agent in these samples after 9 weeks. Biomass production, nutrient and metal bioavailability in the soil were determined, along with their uptake by plants and the resulting remediation factors. Before S application, Cd and Zn mobility was reduced in all the AC, Lig and VC treatments, while Pb mobility was increased only in the Lig1 and VC1 treatments. Upon sulphur application, Fe, Mn, Cd, Pb and Zn mobility was not significantly affected in the C, AC and VC treatments, nor total Cd, Pb and Zn contents in maize shoots. Increased sulphate, Mn, Cd, Pb and Zn mobilities in soil together with related higher total S, Mn, Pb and Zn contents in shoots were observed in investigated treatments in the last sampling period. The highest biomass production and the lowest metal toxicity were seen in the VC treatments. These results were associated with effective metal immobilization and showed the trend of steady release of some nutrients. The highest remediation factors and total elemental content in maize shoots were recorded in the VC treatments. This increased phytoremediation efficiency by 400% for Cd and by 100% for Zn compared to the control. Considering the extreme metal load of the soil, it might be interesting to use highly metal-tolerant plants in future research. Future investigations could also explore the effect of carbonaceous additives on S oxidation, focusing on the specific microorganisms and redox reactions in the soil. In addition, the homogeneous distribution of the S rate in the soil should be considered, as well as longer observation times.