Reduction in growth of saplings of Eucalyptus deglupta resulting from die-back caused by Amblypelta cocophaga China was studied in forestry plantations in the Solomon Islands. At a standard spacing of 10 × 3 m, a comparison was made of tree growth and incidence of A. cocophaga attack on saplings planted in strips 2 m wide cut through secondary forest, with those from which all inter-row vegetation was cut back prior to planting. Initially, a greater proportion of trees was infested on the cleared than the uncleared plots, but at the end of the first year, at which time the attack was at its peak, no differences could be detected. From the fifth to the thirteenth month after planting out, uninfested trees on the cleared plots increased in height at an average rate of 60 cm per month. For every week that at least one A. cocophaga was recorded on a tree, this rate was depressed by 1·3 cm. On the uncleared plots, the average rate of increase of uninfested trees was only 42 cm per month and this was reduced by 1·0 cm for every week that an insect was recorded. By the end of the first year, an estimated 37% (in cleared plots) and 41% (in uncleared plots) of the potential stand was rendered useless for future timber production due to A. cocophaga damage. In terms of total stem volume (not log production), about 1 m3/h out of a potential 3 m3/h was lost in the uncleared plots and 0·2 m3/h out of a potential 0·6 m3/h in the uncleared ones. Whilst clearing did not reduce the incidence of A. cocophaga attack, the advantage of clearing in terms of increased tree growth was considerable.