Field experiments on reduced application doses of Metarhizium anisopliae var. acridum (I 91 609: Zonocerus variegatus strain) were carried out in 1996–1997 in southeastern Benin and central Ghana. The first experiment, which took place in southeastern Benin, consisted of ULV applications of an oil based formulation of M. anisopliae on 1-ha plots of cassava, at the rates of 1 × 1011, 4.65 × 1011 2.5 × 1012 and 1 × 1013 conidia/ha. In a second experiment carried out in central Ghana, the application doses were reduced by increasing the space between spray tracks. Treatments were applied at 5, 10 and 20m track spaces, corresponding to doses of 2.5 × 1012, 1.25 × 1012 and 0.625 × 1012 conidia/ha, respectively. After application, samples of L4–L6 larvae and adults of Z. variegatus were collected using sweep nets and their mortality monitored in small cages. Zonocerus variegatus population density was monitored at three-day intervals over a 21-day period. In southeastern Benin, analysis of survival time, resulting from the first sampling, did not reveal any significant difference between the various dose treatments. In central Ghana, for treatments at different doses and track spaces, there was no difference between 5 and 10m track spaces, corresponding to 2.5 × 1012 and 1.25 × 1012 conidia/ha. But the 1.25 × 1012 conidia/ha treatment at 10 m-track space reduced the application time by half (17 min instead of 34 min).