HPLC-electrospray mass spectrometry was used to identify the phosphorylated sites on a bacterially expressed cystic fibrosis transmembrane conductance regulator (CFTR) fragment containing the first nucleotide binding domain (NBDI) and the regulatory domain (R). Tryptic digests of NBDI-R (CFTR residues 404-830) were analyzed after protein kinase A (PKA) treatment for all possible peptides and phosphopeptides (a total of 1 18 species) containing Ser residues within "high-probability'' PKA consensus sequences: R-R/K-X-SIT, R-X-X-SIT, and R-X-S/T. Three criteria were used to assign phosphorylated sites: ( I ) an 80-Da increase in the predicted average molecular weight of the tryptic peptides; ( 2 ) co-elution with the PO,-ion induced by stepped energy collision; and (3) the relative elution positions of the phosphorylated and unmodified peptides. Ser residues within the eight dibasic sites in the NBDl and R domains (positions 422, 660, 700, 712, 737, 768, 795, and 813) were phosphorylated, a pattern similar to that observed for full-length CFTR. The serine at position 753, which in CFTR is phosphorylated in vivo, was not phosphorylated. The remaining potential PKA sites, Ser4", Ser5I9, Ser670, and Thr7", were not phosphorylated. The "lowprobability" PKA sites (those not containing an Arg residue) were not phosphorylated. The results suggest that isolated domains of CFTR developed useful models for investigating the biochemical and structural effects of phosphorylation within CFTR. The mass spectrometry approach in this study should prove useful for defining phosphorylation sites of CFTR in vitro and in vivo.