This study assessed the relative bioavailability of guanidinoacetic acid (GAA) in cattle. Seven ruminally cannulated Holstein steers (initial body weight of 280 kg) were used in an experiment with a 5 × 5 Latin square design; the two additional steers received a treatment sequence identical to two steers in the Latin square. Treatments were: control (no GAA, water infusion), ruminal infusion of 10 or 20 g/d GAA, and abomasal infusion of 10 or 20 g/d GAA, with all infusions delivered continuously. Periods were 7 d in length, and on day 7, blood and urine samples were collected to determine the concentrations of GAA and its associated metabolites. Plasma creatine concentrations increased linearly (P < 0.01) with GAA infusion to the abomasum and tended to increase linearly (P = 0.06) when GAA was infused ruminally. Urinary creatine concentrations increased linearly with increasing amounts of GAA infused in the abomasum (P < 0.01) and the rumen (P < 0.05). There were no significant effects of GAA infusion to either the abomasum or rumen on plasma or urinary concentrations of GAA. Plasma creatinine concentrations were not affected by GAA infusion to the abomasum or rumen. Urinary creatinine concentrations decreased when GAA was infused abomasally (P < 0.05). Because plasma and urinary creatine concentrations yielded the statistically strongest linear responses, they were selected as the primary response criteria for quantifying ruminal escape of GAA. Calculated by slope-ratio methodology, estimates for the ruminal escape of GAA based on plasma creatine and urinary creatine concentrations were 47% and 49%, respectively. Ruminally infused GAA was about half as effective as abomasally infused GAA in elevating plasma and urinary concentrations of creatine.