Incubation of purified skeletal muscle glycogen synthetase I with trypsin (10 mug/ml) for 15 min decreased the Stokes radius of the enzyme from 68 A to 62 A and the subunit molecular weight from 90,000 to about 73,000. No decrease in the sedimentation coefficient of 13.3 S could be detected. It was calculated that native synthetase I is a tetramer of molecular weight 360,000 to 370,000. Trypsin also catalyzed a decrease in the synthetase activity ratio (minus glucose-6-P to plus glucose-6-P) largely by reducing minus glucose-6-P activity. The magnitude of the trypsin effect on the synthetase activity ratio was very similar to that produced by phosphorylation of synthetase by the cyclic AMP-dependent protein kinase. The activity ratio could be lowered from that characteristic of synthetase I, 0.85, to 0.25 either by trypsin digestion or by incorporation of 1 mol of Pi per mol of synthetase subunit. An activity ratio of less than 0.05 could be obtained in three ways as follows, (a) phosphorylation to 2 Pi/subunit, (b) phosphorylation to 1 Pi/subunit (ratio=0.25) followed by trypsin treatment, (c) trypsin treatment (ratio=0.25) followed by phosphorylation. When trypsinized synthetase was phosphorylated by the catalytic subunit of cyclic AMP-dependent protein kinase, 1 Pi/subunit was incorporated. Trypsin (6 mug/ml) led to a rapid release of about 50% of the radioactivity from 32P-synthetase regardless of whether the enzyme contained 1 or 2 phosphates per subunit. It was concluded that two sites on the enzyme subunit are highly susceptible to phosphorylation catalyzed by the cyclic AMP-dependent protein kinase. The data indicate that the 1st mol of Pi incorporated is distributed about equally between the two sites. A model is proposed to account for these observations. The data further indicate that a peptide containing one site is removed by trypsin. This reduces enzyme activity to the same extent as does phosphorylation of the site in the intact protein.