The magnitude, temporal, and spatial patterns of nitrous oxide (N2O) fluxes in plantations are still largely unknown; however, they are crucial for our understanding and management of global greenhouse gas emissions. The objective of this study was to determine the effects of forest management practices, such as the understory removal and nitrogen (N)-fixing species (Cassia alata [C. alata]) seeding, on soil N2O fluxes in four forest plantations in southern China. Fluxes of N2O were measured in a Eucalyptus urophylla plantation (EUp), an Acacia crassicarpa plantation (ACp), 10 native species-mixed plantation (Tp), and 30 native species-mixed plantation (THp) by a static chamber method from June 2007 to May 2008 in Guangdong province, China. Four forest management treatments, including understory removal and replacement with C. alata (UR+CA), understory removal only (UR), C. alata seeding only (CA), and (4) control without any disturbances (CK), were applied in the four forest plantations. The results showed that N2O fluxes were higher under UR treatment as compared to CK in EUp (16.9 μg m−2 h−1), ACp (16.3 μg m−2 h−1), Tp (14.4 μg m−2 h−1), and THp (14.4 μg m−2 h−1) during the study period. Soil N2O fluxes under CA treatment tended to be enhanced in EUp (18.1 μg m−2 h−1), ACp (18.3 μg m−2 h−1), Tp (19 μg m−2 h−1), and THp (16.6 μg m−2 h−1), having higher values in CA than in CK. There were positive relationships between N2O fluxes and soil temperature (P < 0.01), soil moisture (P < 0.01), and nitrate (NO3)–N concentrations (P < 0.05). Our results indicated that soil NO3–N, soil temperature, and moisture are the primary controlling variables for soil N2O fluxes. The present study improved our understanding of soil N2O fluxes in forest plantations under different management practices.