The kinetics of fat cell adenylyl cyclase were studied, with AMP-P(NH)P and Mn++ or Mg++ as the divalent cation. In general, the reaction times were not linear. In the presence of fluoride or GMP-P(NH)P, the time curves were concave upwards; in other cases (i.e., basal activity, insulin, or isoproterenol), transient rates tended to decrease with time during the assay. Kinetic data were analyzed according to a previously described procedures (Torres et al., 1978b) which isolates two kinetic components: initial and final. With AMP-P(NH)P, kinetic activities were about ten times lower than those for ATP. With Mn++, activities were at least two-times higher than for Mg++. Spontaneous inactivation of adenylyl cyclase was higher in assays containing Mg++ than in those supplemented with Mn++. In the latter case, insulin was able to increase the inactivation rate. Fluoride and isoproterenol both activated adenylyl cyclase in both the initial and final kinetic components; under most of the conditions explored, their effects on the final component appeared to be more dramatic. Assays with GMP-P(NH)P showed inhibited activity in the initial component and increased activity in the final one. When the results obtained with AMP-P(NH)P are compared with those of ATP (Torres et al., 1978b. J. Membrane Biol. 43:000), the following differences were found: (i) in the presence of insulin and Mn++, cyclase inactivation was higher with AMP-P(NH)P than with ATP; (ii) fluoride stimulation of the final component was more marked with ATP than with AMP-P(NH)P; (iii) cyclase stimulation by isoproterenol was slightly higher with the nucleotide analog; and (iv) GMP-P(NH)P stimulation of the final component resulted in higher activity with ATP than with AMP-P(NH)P.