Background: there has been much controversy on the role of perforator veins in the development of chronic venous disease (CVD). This study was designed to determine the duration of flow of lower limb perforator veins (PVs) in relation to their location, diameter and competency status of superficial and deep veins, in healthy volunteers and patients with different grades of CVD. Patients and methods: thirty limbs in 15 symptom-free volunteers and 103 limbs in 75 patients with signs and symptoms of CVD were examined with colour-flow duplex scanning. Superficial, perforator and deep veins were studied in the standing and sitting positions. Flow-velocity characteristics, the number and maximum PV diameter at the deep fascia and subfascially were determined. A PV was considered incompetent when the outward flow lasted > 0.5 s. Results: 581 PVs were found in the patients and 106 in the volunteers. 163 PVs (28%) were incompetent in the first group and none in the latter. The total number of PVs and the number of incompetent PVs per limb increased significantly with the severity of CVD. The mid-calf area had more competent and incompetent PVs in patients (p < 0.01). Mean diameter of incompetent PVs in all the CVD classes was significantly larger than that of competent PVs. Competent PVs tended to be larger with increasing severity of CVD and they were significantly larger in the CVD classes 4 to 6 compared to controls (p < 0.01). Subfascial PV diameter was markedly larger than that at the fascial level (p < 0.001) regardless of the CVD class. A subfascial PV diameter of > 3.9 mm (95% CI 3.4 to 4.4 mm) indicated incompetence. However, the reverse was not true, because about a third of incompetent PVs had a subfascial diameter of < 3.9 mm. Both competent and incompetent PVs were smaller when located at the lower thigh, knee, ankle and anterior aspect of the calf than those found in the rest of the calf and mid-thigh (p = 0.03). Both inward and outward flow was found more often in patients than in controls (70/418 vs. 9/106, p = 0.048). Most incompetent PVs had outward flow alone (126, 77%). PV incompetence was most frequently associated with reflux in superficial veins (120, 74% (p < 0.0001), followed by reflux in both the superficial and deep veins (34, 21%) and reflux in the deep veins alone (9, 5%). The mean duration of outward flow was markedly longer in the presence of both superficial and deep vein reflux compared to superficial (p < 0.001) or deep vein reflux alone (p < 0.0001). Conclusions: the number of incompetent PVs and the diameter of both competent and incompetent PV increases with the severity of CVD. Bidirectional PV flow is more common in patients than in normal volunteers, while 77% of the incompetent PVs have outward flow alone. PV incompetence is most often associated with reflux in the superficial veins, indicating that deep venous reflux is rarely the primary cause of PV insufficiency.