Purpose: To implement and characterize a fluorine-19 (F-19) magnetic resonance imaging (MRI) technique and to test the hypothesis that the F-19 MRI signal in steady state after intravenous injection of a perfluoro-15-crown-5 ether (PCE) emulsion may be exploited for angiography in a pre-clinical in vivo animal study. Materials and Methods: In vitro at 9.4T, the detection limit of the PCE emulsion at a scan time of 10 min/slice was determined, after which the T-1 and T-2 of PCE in venous blood were measured. Permission from the local animal use committee was obtained for all animal experiments. 12 mu l/g of PCE emulsion was intravenously injected in 11 mice. Gradient echo H-1 and F-19 images were obtained at identical anatomical levels. Signal-to-noise (SNR) and contrast-to-noise (CNR) ratios were determined for 33 vessels in both the F-19 and H-1 images, which was followed by vessel tracking to determine the vessel conspicuity for both modalities. Results: In vitro, the detection limit was similar to 400 mu M, while the F-19 T-1 and T-2 were 1350 +/- 40 and 25 +/- 2 ms. The F-19 MR angiograms selectively visualized the vasculature ( and the liver parenchyma over time) while precisely coregistering with the H-1 images. Due to the lower SNR of F-19 compared to H-1 (17 +/- 8 vs. 83 +/- 49, p<0.001), the F-19 CNR was also lower at 15 +/- 8 vs. 52 +/- 35 (p<0.001). Vessel tracking demonstrated a significantly higher vessel sharpness in the F-19 images (66 +/- 11 vs. 56 +/- 12, p=0.002). Conclusion: F-19 magnetic resonance angiography of intravenously administered perfluorocarbon emulsions is feasible for a selective and exclusive visualization of the vasculature in vivo.