The rehabilitation of vegetation on structurally crusted soils by triggering termite activity through mulch was studied on three soil types in northern Burkina Faso, West Africa. A split-plot design was used in a fenced environment for the experiment. Insecticide (Dieldrin) was used at a rate of 500 g a.i. (active ingredient)/ha to create nontermite and termite plots. Three mulch types consisting of straw (Pennisetum pedicellatum), woody material (Pterocarpus lucens), and a composite mulch (straw and woody material) applied at a rate of 3, 6, and 4 tons/ha, respectively, were used to trigger termite activity. The grasses and woody species on the plots were surveyed. Nontermite plots responded weakly to mulch treatments, but even in the first year vegetation established on termite + mulch plots. Termite activity resulted in the increase of plant cover, plant species number, phytomass production, and rainfall use efficiency. Infiltrated water use efficiency and plant diversity were not statistically different among treatments during the first 2 years but were in the third. Woody species established only on termite plots. The three types of mulch plots showed greater vegetation development than bare plots, which remained bare throughout the experiment. Analysis of the termite and mulch interaction indicated that mulch plots without termites did not perform better than bare plots, especially in the case of woody plant regeneration. Vegetation rehabilitation was best with composite and straw mulches with termites, followed by woody mulch with termites; it was worst on bare plots.