Purpose Pro-angiogenic factors are positively associated with breast tumor staging and poorer prognosis, but their role in the etiology of breast cancer has not been assessed. Methods: We measured serum levels of the pro-angiogenic vascular endothelial growth factor A (VEGF), and placental growth factor (PlGF) and anti-angiogenic soluble fms-like tyrosine kinase-1 (sFlt-1) in 352 incident breast cancer cases [mean age at diagnosis 67 (range 55–83)] and 352 non-cases in the prostate, lung, colorectal, and ovarian screening trial (women enrolled 1993–2001, followed through 2005) matched on age and date of enrollment. Cases were followed on average 4.2 years from blood draw to diagnosis, range 3.9–12.8 years; 53 % were estrogen receptor positive/progesterone receptor positive (ER+/PR+), and 13 % were ER−/PR−. Quartile-specific hazard ratios (HR) and 95 % confidence intervals (CI) were estimated using weighted Cox proportional hazards regression models adjusted for known breast cancer risk factors. An ordinal variable for the angiogenic markers was used to test for trend in the HR. Results: Comparing the highest to lowest quartile, multivariable HR were 0.90 for VEGF (95 % CI 0.33–2.43, p trend = 0.88), 1.38 for sFlt-1 (95 % CI 0.63–3.04, p trend = 0.63), and 0.62 for PlGF (95 % CI 0.19–2.00, p trend = 0.73). Risk patterns were not altered when all angiogenic markers were included in the model simultaneously, or by restricting analyses to invasive breast cancers, to cases diagnosed two or more years after blood collection or to ER+ tumors. Conclusions: There was no evidence of an increased breast cancer risk associated with circulating levels of pro-angiogenic markers VEGF and PlGF or a reduced risk with circulating levels of anti-angiogenic marker sFlt-1. Electronic supplementary material The online version of this article (doi:10.1007/s10552-016-0779-5) contains supplementary material, which is available to authorized users.