Summary Ecological restoration activities, including reforestation, often involve the use of herbicides for the removal of weedy plant cover. Little is known, however, about the effects of herbicides on assemblages of non-target organisms that colonize restored patches. We describe a field experiment to investigate effects of glyphosate herbicide (Roundup® Biactive™) on rainforest-associated soil- and litter-dwelling macro-arthropods. Our experimental protocol differed in two ways from other ecotoxicological studies of herbicides. First, we applied herbicide at a rate considerably greater than the manufacturer's recommended maximum in order to simulate worst-case scenarios that may occur in the practice of forest restoration. Second, our field experiment was carried out under dense canopy cover with sparse understorey vegetation, so that indirect impacts caused by the loss of existing vegetation were eliminated. Paired herbicide-treated and control plots were created within five rainforest remnants on the Maleny plateau of subtropical eastern Australia. Macro-arthropods were collected using litter extraction before, approximately 3 days after, and 3 months after herbicide application. Responses of arthropods were analysed at two levels of taxonomic resolution: ‘coarse’ arthropods (arthropods sorted to Order/Class), and ant species. Our results suggest that the use of glyphosate herbicide formulated as Roundup® Biactive™ is suitable for the control of unwanted plants in rainforest restoration sites as it appears to have minimal impact on assemblages of soil and litter macro-arthropods or at least those typical of intact rainforest.