Purpose Disturbed ocular haemodynamics are discussed to contribute to the pathogenesis of glaucoma. Up to now there is no method available allowing direct determination of blood flow, which is the most relevant dimension for studies on haemodynamics. In this study, volumetric colour Doppler imaging (vCDI) is evaluated systematically in glaucoma patients. Methods A Siemens Elegra ultrasound set-up with a linear 7.5 MHz probe was used for all CDI measurements. For vCDI, the crosssectional area of a vessel and the flow velocity is determined. From both these parameters blood flow can be calculated. Ocular pulse amplitude (OPA) was assessed by the method of Langham using a pneumatic applanation tonometer. Results (1) Velocity measurements using CDI in the ophthalmic artery and central retinal artery were highly reproducible (n ¼ 20). In contrast, reproducibility of vCDI measurements was low (n ¼ 20). Reproducibility improved if five vCDI measures were averaged. (2) Results from two different CDI-operators did not differ regarding the velocity measurements, but there was a difference in vCDI measurements (n ¼ 20). (3) Results from vCDI did not correlate with measurements of OPA in 69 patients. (4) In 15 patients, vCDI failed to detect changes of ocular perfusion induced by the application of dorzolamide. Conclusion vCDI is not applicable in ophthalmology at present.