The order of first limiting AA in late gestation was determined in 14 sows (6 gilts, 8 2nd parity sows) using the AA deletion technique. A positive control (PC) diet was formulated to contain all essential AA at 160% of current recommendations (NRC 1998). Lysine, threonine, tryptophan and branchedchain AA were each removed from the PC diet to 60% of current recommendations (PC-Lys, PC-Thr, PC-Trp, PC-BCAA, respectively). Sows received each of the diets in random order. Plasma urea nitrogen, indicator AA oxidation and protein turnover (L[1-13C]phenylalanine as tracer AA) were used to evaluate the response to dietary AA deletion from 90-111 d of gestation. Data was analyzed using a Mixed model with sow as random variable. Parity was tested as a covariate but was not significant. Fasted plasma urea nitrogen (mg/dL) was highest (P < 0.09) for the PC-Lys (11.77) diet compared to the PC (8.89) diet followed by the PC-BCAA (10.81) and PCTrp (10.74) diets. Indicator AA oxidation (%) tended to be higher (P < 0.15) for the PC-Thr (18.8) and PC-Trp (18.4) diets compared to the PC (16.4) diet. Tracer flux and protein breakdown (μmol/kg body wt per h, P < 0.05) were higher on the PC-Trp (91.57 and 45.08, respectively) diet compared to the PC (73.72 and 27.97, respectively) diet. The first limiting AA in late gestation may be Thr or Trp rather than Lys but appears to be dependent on the response parameter. Changes in the order of first limiting AA in late gestation in sows means that the traditional Lys to AA ratio used in diet formulations, based on growing pig data, may not apply to pregnant sows.