The type-1 protein phosphatase associated with hepatic microsomes has been distinguished from the glycogen-bound enzyme in five ways. (1) The phosphorylase phosphatase/synthase phosphatase activity ratio of the microsomal enzyme (measured using muscle phosphorylase a and glycogen synthase (labelled in sites-3) as substrates) was 50-fold higher than that of the glycogen-bound enzyme. (2) The microsomal enzyme had a greater sensitivity to inhibitors-1 and 2. (3) Release of the catalytic subunit from the microsomal type-1 phosphatase by tryptic digestion was accompanied by a 2-fold increase in synthase phosphatase activity, whereas release of the catalytic subunit from the glycogen-bound enzyme decreased synthase phsophatase activity by 60%. (4) 95% of the synthase phosphatase activity was released from the microsomes with 0.3 M NaCl, whereas little activity could be released from the glycogen fraction with salt. (5) The type-1 phosphatase separated from glycogen by anion-exchange chromatography could be rebound to glycogen, whereas the microsomal enzyme (separated from the microsomes by the same procedure, or by extraction with NaCl) could not. These findings indicate that the synthase phosphatase activity of the microsomal enzyme is not explained by contamination with glycogen-bound enzyme. The microsomal and glycogen-associated enzymes may contain a common catalytic subunit complexed to microsomal and glycogen-binding subunits, respectively. Thiophosphorylase a was a potent inhibitor of the dephosphorylation of ribosomal protein S6, HMG-CoA reductase and glycogen synthase, by the glycogen-associated type-1 protein phosphatase. By contrast, thiophosphorylase a did not inhibit the dephosphorylation of S6 or HMG-CoA reductase by the microsomal enzyme, although the dephosphorylation of glycogen synthase was inhibited. The I 50 for inhibition of synthase phosphatase activity by thiophosphorylase a catalysed by either the glycogen-associated or microsomal type-1 phosphatases, or for inhibition of S6 phosphatase activity catalysed by the glycogen-associated enzyme, was decreased 20-fold to 5–10 nM in the presence of glycogen. The results suggest that the physiologically relevant inhibitor of the glycogen-associated type-1 phosphatase is the phosphorylase a-glycogen complex, and that inhibition of the microsomal type-1 phosphatase by phosphorylase a is unlikely to play a role in the hormonal control of cholesterol or protein synthesis. Protein phosphatase-1 appears to be the principal S6 phosphatase in mammalian liver acting on the serine residues phosphorylated by cyclic AMP-dependent protein kinase.