The copia insertion responsible for the wa mutation is 3' to the white promotor and in the same transcriptional orientation as white. First, we have analyzed the effects of the w a copia insertion on levels of polyadenylated white transcripts and find large, developmentally programmed effects. Second, we have isolated and sequenced an LTR-excision event involving the copia insertion at wa. This represents the first documented case of an LTR-excision event in Drosophila. This single copia LTR has developmentally programmed effects on white transcript levels qualitatively similar to the intact copia element. Third, we have characterized the structures of white transcripts from wa. We find polyadenylated white transcripts apparently having 3' termini in or near the 3' LTR of the wa copia insertion, as has been reported in limited studies of wa transcription in adults by others. These earlier studies also revealed wa transcripts apparently corresponding to polyadenylated terminus formation in the 5' LTR of the copia transposon; however, our more detailed studies reveal that these transcripts probably have other origins and that little, if any, polyadenylated terminus formation for white transcripts occurs in the 5' LTR of the wa copia insertion. Moreover, we find no polyadenylated terminus formation for white transcripts occurring in the single LTR of the wa LTR-excision product. Fourth, we find that each of three mutant alleles at su(wa) produces elevated levels of several classes of RNAs apparently corresponding to transcriptional readthrough of the wa copia transposon. Elevated levels of one presumptive readthrough transcript were observed previously in one su(w a)1 mutant strain. Fifth, we have confirmed the existence of a transcript initiated in the 3' LTR of the wa copia insertion and find the levels of this transcript to be strongly influenced by developmental stage and genetic background. Lastly, we have analyzed white transcripts produced by the whd81b11 allele, which carries an insertion of copia in the opposite transcriptional orientation and in a different position than the wa copia insertion. In contrast to the wa copia insertion allele, the whd81b11 allele produces polyadenylated white transcript levels very similar to the w+ case at the stages examined. Moreover, the whd81b11 copia element apparently produced polyadenylated terminus formation in white transcripts and we observe no effect of the allelic state of su(wa) on apparent readthrough of this stop site. We discuss some possible implications of our results for the properties of retrotransposons as developmentally complex insertional mutagens and for the functional organization of the copia transcription unit.