Six mature Yorkshire X Landrace sows were randomly assigned to a 6 X 6 Latin-square experiment on d 3 of lactation to determine the phenylalanine (Phe) requirement in the presence of excess tyrosine (Tyr). A control diet of corn sugar, cornstarch, whey, L-glutamic acid, solka floc, soybean oil, amino acids, minerals and vitamins was supplemented with L-Phe to provide .175, .250, .325, .400, .475 and .550% Phe and .47% Tyr. Each diet was fed to a maximum of 5.5 kg/d within each of six 7-d periods. Sow milk yield and average pig weight gain decreased (cubic, P less than .01) with increasing time period. Feed intake and average pig weight gain were depressed (quadratic, P less than .05) on the lowest Phe diet, but milk yield was not significantly affected. Daily nitrogen (N) intake was depressed (quadratic, P less than .01) when sows consumed the .175% Phe diet, but percent fecal N was not affected. Percent urine N decreased (quadratic, P less than .001) and percent N retained (without milk) increased with increasing Phe to .288 and .296%, respectively. Daily secretion of milk N as a percentage of N intake was not affected by diet. Percent overall N balance (with milk) increased with increasing Phe to .307%. Urea N decreased with increasing dietary Phe in both plasma (linear, P less than .01) and urine (quadratic, P less than .001) to a breakpoint at .285% Phe. A sharp rise (quadratic, P less than .01) in plasma Phe occurred at .299% dietary Phe. Plasma Tyr increased (quadratic, P less than .05) as dietary Phe increased, but no clear inflection point was obtained.(ABSTRACT TRUNCATED AT 250 WORDS)