A recurrent selection breeding technique was used to reduce leaf feeding damage by 1st-generation European corn borer (ECB), Ostrinia nubilalis Hubner, and to increase DIMBOA (2,4-dihydroxy-7-methoxy-(2H)-1,4-benzoxazin-3(4H)-one) content in a synthetic maize, Zea mays L., cultivar (BS1); chemical analysis was for MBOA (6-methoxybenzoxazolinone). TLC plate ratings (intensity of MBOA on TLC plates) were also determined. Cycle populations (C0, C1I, C2I, C3I, C1D, C2D, C3D), population crosses (C1I x C1D, C2I x C2D, C3I x C3D) and checks were evaluated. the C1I, C2I, and C3I populations were selected by the insect, whereas the C1D, C2D, and C3D populations were selected by chemical analysis for DIMBOA. Leaf feeding damage caused by the insect, MBOA concentration in whorl leaf tissue, and TLC plate ratings of MBOA intensity on TLC plates were all improved with recurrent selection procedures when selection was made by the insect and also when selection was made by chemical analysis for MBOA in leaf tissue. The gene action for these traits seems to be primarily additive. Each method of selection accumulated about the same number of genes for resistance. A total of 100 S(,1) lines from each of the C0, C3I, and C3D populations were evaluated for leaf feeding resistance by 1st-generation ECBs. Mean leaf-feeding ratings of the S(,1) lines from C0, C3I, and C3D populations were 4.3, 2.8, and 2.5, respectively. Means of the C0 S(,1)s were significantly greater than were the means of the C3I and C3D S(,1)s. The mean of the C3I S(,1)s was greater than was the mean of the C3D S(,1)s. These results are consistent with the results of the C3 populations per se. The cycle populations and population crosses were susceptible to sheath-collar feeding and stalk feeding by 2nd-generation ECBs;DIMBOA (mg MBOA/g dried leaf tissue) content in leaf tissue at five plant heights (15, 30, 41, 60, 83 cm) and DIMBOA content in sheath-collar tissue at two growth stages (anthesis and 3 weeks past anthesis) were determined for C0, C3I, and C3D maize populations plus inbreds CI31A and WF9. CI31A contained more MBOA and WF9 contained less MBOA than did all other genotypes. The C0 population contained less MBOA than did the C3I and C3D populations. Plants 15-cm tall contained more MBOA than did plants 83-cm tall. TLC plate ratings were significantly different among genotypes but not among plant heights.