Widely used phthalates, especially di-(2-ethylhexyl) phthalate (DEHP), increase breast cancer risk in experimental animals and humans, but long-term follow-up evidence of its human breast carcinogenicity remains inconclusive. This nested case–control study included 119 invasive breast cancer cases and 245 matched controls from a longitudinal cohort of 11,923 women recruited in 1991–1992 and followed to 2010 in Taiwan. Urine samples at baseline and follow-up visit were tested for 11 metabolites of seven phthalates using LC-ESI-MS/MS. DEHP metabolism susceptibility was evaluated by the percentage of mono-2-ethylhexyl phthalate (MEHP%) in the sum of five DEHP metabolites (∑DEHP). Odds ratios (ORs) with 95% CI from conditional logistic regression were used to examine risk predictors. DEHP was the only phthalate significantly associated with breast cancer risk. Risk increased significantly with elevated urinary levels of ∑DEHP (> 0.381 μmol/g creatinine, OR = 1.71, 95% CI = 1.02 to 2.43), MEHP (> 0.022 μmol/g creatinine, OR = 1.87, 95% CI = 1.07 to 3.25), and MEHP% (> 6.7%, OR = 1.65, 95% CI = 0.96 to 2.82). Elevated ∑DEHP and MEHP% combined with early menarche (≤ 14 years) was associated with further increased risk (OR = 7.52, 95% CI = 2.68 to 21.05). The intraclass correlation coefficient between paired baseline and follow-up samples of 152 women was 0.06 for ∑DEHP and 0.31 for MEHP%. High DEHP exposure, high MEHP%, and early menarche were associated with increased breast cancer risk. MEHP% was a better biomarker for DEHP metabolism.