The mechanism of GTPase-activating protein (GAP) activation of p21ras GTP hydrolysis has been investigated by measuring the kinetics of release of Pi during the hydrolysis. The measurement uses a continuous spectroscopic assay for Pi, based on a guanosine analogue, 2-amino-6-mercapto-7-methylpurine ribonucleoside, as substrate for purine nucleoside phosphorylase [Webb, M.R. (1992) Proc. Natl. Acad. Sci. U.S.A. 89, 4884-4887]. This phosphorolysis gives an absorbance increase at 360 nm, so that when the reaction is coupled to GTP hydrolysis, the change in absorbance gives the total amount of Pi released from the p21ras. The rate of the absorbance increase gives the GTPase activity. This provides a non-radioactive method of determining p21ras concentration and GAP activity. It was used to determine the interaction of GAP with wild-type p21ras and two mutants (Leu-61/Ser-186 and Asp-12), all in the GTP (or guanosine 5'-[ beta gamma-imido]triphosphate) form. The Leu-61/Ser-186 mutant binds 10-fold tighter than does the wild-type protein. The Asp-12 mutant binds to GAP with the same affinity as the wild-type protein. A novel GTPase activity was characterized whereby the EDTA-induced nucleotide release and GAP-activated cleavage of bound GTP leads to steady-state turnover of GTP hydrolysis. An assay for GAP is described based on this activity.