Differential intra-and inter-row weed management can be a means to reduce herbicide use in sugarcane. A field experiment was conducted in 2011 in La Reunion Island to assess inter-row weed competition. Four inter-row weed competition treatments for a duration of one (T1), two (T2), three (T3) and four (T4) months after planting were imposed in a randomised complete block design with 5 replications; treatment plots were paired with non-weeded inter-row control plots. Row spacing was 1.5 m and space allocated to intra-row and inter-row was equally divided. All intra-rows where kept weed-free all season long. Weeds covered 100% of the inter-rows by three months after planting. Sugarcane grew far above the weed canopy, completely closing the canopy between the rows for all treatments. Overall, sugarcane tillering and production was significantly reduced by weed competition. Sugarcane yield decreased by 13 t/ha/month of early competition. However, the control yield, expected to be lower than T4, was similar to the T3 treatment. This suggests a late facilitation effect of the inter-rows vegetation in control plots which partially compensates for its early competition effect. This may be due to the particular flora of our experiment, with high populations of N-fixing weeds and broomweeds. Obviously, additional research is needed to corroborate this singular result.