In summer of 1989 we conducted the world's first field test of cotton genetically engineered to contain a gene from Bacillus thurigiensis which codes for the delta endotoxin. The strains were genetically engineered by Agracetus located in Middleton, Wisconsin and the field plots and laboratory growth studies were conducted at Mississippi State, Mississippi. We evaluated two insertions and had a homozygous and a heterozyqous strain of each insertion. The parental line was 'Coker 312'. The adapted cultivar used for comparison was 'DES 119'. Plots were grown in a randomized block experiment with 6 replications with a split plot design. Whole plots were two treatments: (1) complete protection from insects, and (2) artificial infestation with first instar tobacco budworm, Heliothis virescens, larvae. The genetically engineered strains were not sufficiently resistant to control the tobacco budworm at the high levels with which we infested. Laboratory growth studies were conducted on excised plant tissue. Using old leaves we found a small reduction in weight after 13 days on two strains. We found no reduction in growth of larvae on new leaves, squares or anthers. Yield of the genetically engineered strains was very good. One insertion yielded 1418 pounds of lint per acre and was not significantly different from the yield of Coker 312 or DES 119. The other insertion yielded 1293 pounds of lint per acre which was significantly lower than DES 119 and Coker 312. One insertion had a lint percent that was significantly higher than the parental strain, Coker 312. The increase in lint percentage was about 1%. Both insertions produced bolls significantly smaller than Coker 312. Boll size was reduced about 500 mg in one insertion and 125 mg in the other insertion. One insertion significantly increased 2.5% span length from 31.4 for Coker 312 to 32.2 for the genetically engineered strain. No other agronomic effects were noted. The genetically engineered strains performed very much like normal cotton strains in this test. They were as good as, but no better than the parental strain, Coker 312, in controlling the tobacco budworm infestations.