Two experiments were conducted to determine the lysine and threonine requirements of gestating sows. In the first experiment, four levels of lysine (0.34,0.42,0.48, and 0.56% crude lysine, and 0.24,0.31, 0.38, and 0.45% standardized ileal digestible lysine) were compared in eight multiparous Large White sows. Each sow received successively the four diets according to a Latin-square experimental design. Nitrogen balance was measured over 11 d after a 10-d period of adaptation to the experimental diet. In the second experiment, four threonine/lysine ratios (0.63, 0.73, 0.80, and 0.89 on a crude basis and 0.61, 0.71, 0.77, and 0.87 on a standardized ileal digestible amino acid basis) were compared in 16 multiparous sows, according to a Latin-square experimental design. The standardized ileal digestibility of amino acids in the experimental diets was determined with ileo-rectal anastomized growing pigs. In the first experiment, nitrogen retention was affected by lysine supply (linear, P < 0.001; quadratic, P < 0.04). Nitrogen retention was lowest for treatment 1 (8.0 g/ d) and highest for treatments 3 and 4 that did not differ. Nitrogen retention plateaued at 14.7 g/d in sows consuming 10.5 g/d of digestible lysine. The maintenance requirement for digestible lysine was calculated to be 27 mg/kg BW(0.75) with an efficiency of utilization of digestible lysine above maintenance at 59%. In the second experiment, nitrogen retention was affected (P < 0.03) by the threonine:lysine ratio. It was lower for the lowest threonine:lysine ratio (0.63) than for the other three treatments that did not differ among each other. These results indicate that the optimal standardized digestible threonine:lysine ratio appears to be about 0.71 for multiparous gestating sows.