Excessive fertilization and accelerated erosion, due to intensive agriculture, have resulted in a large nitrogen (N) and phosphorus (P) exports from farmlands to rivers and lakes, contributing substantially to eutrophication in many regions. But many studies mainly emphasized the influence of fertilizer overuse, neglecting the contribution of accelerated erosion to diffuse N and P pollution. Thus, the objectives of this study were to determine the seasonal changes of sediment and particulate N (PN) and P (PP) entering the river by surface runoff from core sugarcane growing watershed, and clarify their influencing factors. Surface runoff, sediment and associated nutrients were measured in summer and autumn at the outlets of three sub-watersheds under rainfall events, in Nala watershed, southern subtropics of China. The total sediment, PN and PP loads in the entire seasons were 30.9 t ha−1, 41.1 kg ha−1 and 14.7 kg ha−1, respectively. The seasonal exports from the entire watershed in early summer, late summer, early autumn and late autumn were respectively, 8.5, 10.5, 11.6 and 0.3 t ha−1 for sediment, 15.3, 12.4, 12.9, and 0.4 kg ha−1 for PN, and 3.4, 5.3, 5.9, and 0.1 kg ha−1 for PP. The largest sediment and PP exports in late autumn coincided with the rainfall peak (208.8 mm), but the highest PN export in early summer was due to sugarcane establishment and basal fertilizer application in the watershed. Particulate N and P exports across the seasons were equivalent to 14 % and 64 % of applied N (296.4 kg ha−1 yr−1) and P (22.8 kg ha−1 yr−1) fertilizers, respectively, but PP exports were highest when no fertilizer and 31 % of P fertilizer were applied. Surface runoff had significant positive relationships with exported sediment and particulate nutrient loads, explaining 41–72 % of the seasonal variations in sediment, PN and PP. Ground covered by sugarcane plants was inversely correlated with sediment, PN and PP exports, and explained 15–40 % of the seasonal variations. Applied N fertilizer only had a positive relationship with PN, whereas no significant relationship (P > 0.05) exists between P fertilizer and PP. The contribution of the sub-watersheds to the export of sediment and particulate nutrients only varied spatially with slope gradient. Thus, our results imply that surface runoff played a dominant role in particulate N and P exports from the sugarcane watershed, but the seasonal variations of particulate N and P exports were dependent on a combined effect of surface runoff and fertilization.