A pot experiment in Kenya examined the effects of mycorrhizal and rhizobial inoculation on the growth and symbiont infection of Acacia tortilis seedlings in unsterile soil, in comparison with sterile and unsterile soil controls. Three mixed, arbuscular mycorrhizal inocula, originating from A. Tortilis, Terminalia brownii, T. Spinosa and Prosopis juliflora (produced under nursery conditions), were used to inoculate the seedlings, which also received a standard, mixed, rhizobial inoculum. Mycorrhizal infection and plant growth were significantly affected by experimental treatment (p<0.001). Some mycorrhizal infection (presumably with indigenous soil-borne inoculum) occurred in the unsterile soil control, but not in the sterile soil control, indicating that transfer of inoculum between pots in the experiment did not occur. Mycorrhizal formation was significantly more extensive in inoculated plants, which also had better shoot and root growth. 24 weeks after inoculation, plants which had received inoculum of A. Tortilis origin were 105 mm tall and their roots were 56% mycorrhizal, whereas the controls most representative of normal nursery treatment were only 77 mm tall and 29% mycorrhizal. Nodulation was erratic and poor in all treatments. The results suggest that mycorrhizal inoculation can improve both mycorrhizal infection and growth of tree seedlings in unsterile nursery soil. The methods of culture and inoculation described here could easily be adopted by nurseries at Little cost.