Expression of the pyrC gene in Escherichia coli K-12 is regulated by a translational control mechanism in which CTP (and perhaps GTP) pool sizes determine the selection of alternative transcriptional start sites at the pyrC promoter. High CTP levels cause transcription to start primarily at a site that directs the synthesis of untranslatable pyrC transcripts. These transcripts form a hairpin at their 5' ends that blocks ribosome binding to the Shine-Dalgarno (SD) sequence. The pyrC ribosome binding site is unusual in that it contains two potential SD sequences, designated SD1 and SD2, which are located 11 and 4 nucleotides upstream of the translational initiation codon, respectively. In this study, we examined the functions of these two SD sequences in translational initiation. Mutations that inactivate either SD1 or SD2 were constructed and incorporated separately into a pyrC::lacZ protein fusion. The effects of the mutations on pyrC::lacZ expression, regulation, and transcript levels were determined. The results indicate that SD1 is the only functional pyrC SD sequence. The SD2 mutation did cause a small reduction in expression, but this effect appeared to be due to a decrease in transcript stability. In addition, we constructed a mutation that introduces a long spacer region between the hairpin at the 5' end of the pyrC transcript and a new pyrC SD sequence. As predicted by the model for translational control, this mutation caused constitutive expression of a pyrC::lacZ protein fusion.