Purpose There is considerable interest in finding a costeffective method of site preparation that effectively controls weeds during planting and further reduces the need for recurring herbicide applications. In this study, two weed control methods, herbicide and scalping, were examined. Both methods may have implications for soil organic matter (SOM) dynamics and nitrogen (N) which could consequently affect plant survival and vegetation establishment. This study aimed to investigate the dynamics of SOM, carbon (C) and N pools under site manipulation practices and the associated early plant survival and growth in tropical Australia. Materials and methods A field trial was established in Central Queensland to examine the recovery of SOM, C and N pools following scalping and the alternative site preparation technique of sequential herbicide application. Both were contrasted with control plots which received neither treatment. Plant survival and growth were also monitored to improve our understanding of plant response to site preparation practices. Results and discussion Scalped plots showed significantly lower values for labile C and N pools compared with the herbicide treatment and control. Generally, there was no significant difference between the herbicide and control for any of the parameters tested. Our observation indicated that herbicide application was significantly less effective than scalping to control weeds. A general decline in SOM parameters was observed in all the plots, including the control during the trial. Drought conditions were considered to be a major factor in the overall decline of SOM. Despite removal of the top soil, there was no significant difference in plant survival between herbicide and scalped areas (81% and 79% survival, respectively). Plant growth was not affected by the treatments in the first 6 months when weed competition in the herbicide areas and low nutrient availability in the scalped plots would have been significant factors in controlling growth rates. However, plants in the herbicide areas, irrespective of species, showed stronger growth than those in the scalped plots at week 61 when they had outgrown the weed competition. It is likely that differences in plant growth response to treatments will become negligible over time. Conclusions Top soil removal was more effective than the use of herbicide in the long-term control of weeds. However, lower SOM and N availability in the scalped areas did not affect plant survival rates when compared with that of the herbicide areas. Whilst the preservation of soil organic matter is considered to be vital in short-term cropping systems, our results indicate that this is not the case for woody vegetation establishment and, in terms of cost and reduction of chemical use, removal of the organically rich top soil, with its accompanying seed burden may be both practical and desirable.