The time course and magnitude of the effects of product loss on the measurement of local cerebral glucose utilization (LCGU) by the 2-[14C]deoxyglucose (DG) method were studied by determination of LCGU in 38 rats with 25-120 min experimental periods after a [14C]DG pulse and in 45 rats with experimental periods of 2.5-120 min during which arterial plasma [14C]DG concentrations (C*P) were maintained constant. LCGU was calculated by the operational equation, which assumes no product loss, with the original set of rate constants and with a new set redetermined in the rats used in the present study; in each case the rate constants were those specific to the structure. Data on local tissue 14C concentrations and C*P were also plotted according to the multiple time/graphic evaluation technique ("Patlak Plot"). The results show that with both pulse and constant arterial inputs of [14C]DG the influence of the rate constants is critical early after onset of tracer administration but diminishes with time and becomes relatively minor by 30 min. After a [14C]DG pulse calculated LCGU remains constant between 25 and 45 min, indicating a negligible effect of product loss during that period; at 60 min it begins to fall and declines progressively with increasing time, indicating that product loss has become significant. When C*P is maintained constant, calculated LCGU does not change significantly over the full 120 min. The "Patlak Plots" reinforced the conclusions drawn from the time courses of calculated LCGU; evidence for loss of product was undetectable for at least 45 min after a pulse of [14C]DG and for at least 60 min after onset of a constant arterial input of [14C]DG.