Abstract Harrowing and fertilisation are common practices at middle rotation in Eucalyptus globulus Labill. plantations in Central Portugal. In order to clarify the effects of such practices on understory vegetation and timber production, a field trial was installed in a 5-year-old first rotation eucalyptus plantation, in a region with mixed oceanic and Mediterranean climatic influences. Four treatments that involved harrowing (H), fertilisation (F), harrowing and fertilisation (HF), and control (C) were tested in the study. The treatments were replicated four times and arranged in a simple completely randomised design. Vegetation surveys were performed by the quadrat method in the 3 years following treatments and by the line interception method in the 7th and 8th years. Samples of understory biomass were collected, oven dried and weighed. In treatments with harrowing, the understory vegetation consistently had lower number of species, less plant cover, species diversity, and biomass than the other treatments. The mean total number of species only once reached 10 in H or HF plots, and was always greater than 12 in C and F plots in the first 3 years, but decreased in the 7th and 8th years. In the first 3 years, the understory biomass averaged 30–60 g m−2 in the F and C plots, and never exceeded 13 g m−2 in treatments with harrowing, which corresponded with the proportion of soil coverage by understory vegetation (4–12% in H and HF, and 38–62% in F and C plots). In the 7th and 8th years, differences in the understory biomass were less important, but the control plots consistently had the largest understory biomass. The influence of treatments in timber production was not statistically significant at the end of rotation.