Despite chronic dialysis treatment, patients with end stage renal disease undergoing maintenance haemodialysis (HD) remain at a substantially increased risk of morbidity. Previous reports using Doppler ultrasound (DU) during HD have revealed microembolic signals (ME) in the venous circulation. In vitro studies confirm the emergence of microbubbles of air that may pass the security system of the HD circuit without triggering the alarm. The aim of this thesis was to elucidate the presence of ME during HD and examine methods that might reduce exposure to ME in vivo. The first study utilized DU to verify the presence of ME in 40 patients during standard HD. Investigation within 30 minutes after the start of HD and just before the end of session revealed the presence of ME in the venous blood line during both phases. The air trap did not alert for the presence of ME. This indicated that ME may pass into the patient during the entire HD run. Study 2 analyzed the presence of ME prior to start and during HD when measured at the AV-access and also carotid artery. A total of 54 patients were examined using DU as the investigative technique. ME increased significantly after start of HD in the AV-access, but also at the carotid artery site. These data indicated that ME can enter the body and even pass the lung barrier. The question arose if microbubbles of air are resorbed or may cause ischemic lesions in organs such as the brain. Study 3 examined whether the amount of ME detected in the AV-access would change by using either a high or a low blood level in the venous air trap/chamber. This was a prospective, randomized and double-blind study of 20 HD patients who were their own controls. After 30 min of standard HD, measurement of ME with DU was performed for two minutes. The chamber setting was changed and after another 30 minutes a new recording was carried out for two minutes. Data showed that setting a high blood level significantly reduced the extent of ME that entered the patient. The results also indicated that ME consisted mainly of microbubbles. In study 4, twenty patients were randomized in a cross-over setting of HD. Three options were used: a wet-stored dialyzer with high blood level (WH) and a dry-stored dialyzer using either a high (DH) or a low (DL) blood level in the venous chamber. The exposure of ME, detected by DU, was least when using mode WF, more with mode DH, and most with mode DL. There was a correlation between higher blood flow and more extensive exposure to ME. Study 5 was an autopsy study of a chronic HD patient with the aim of searching for microbubbles deposited in organs. Microbubbles of gas were verified in the vessels of the lungs, brain and heart. By using a fluorescent stain of anti-fibrinogen it was verified that the microbubbles were covered by clots that had to be preformed before death occurred. This indicated that air microbubbles are not completely absorbed and could result in embolic deposition in the organs of HD patients. In conclusion, these in vivo studies showed that ME pass the air trap without inducing an alarm and enter the venous blood line of the patient. The data confirmed the presence of ME in the AV-access and also in the carotid artery. Autopsy data of a deceased HD patient demonstrated the presence of microbubbles in the capillaries of the lungs, but also in the systemic circulation such as in the brain and the heart. A high blood level in the venous chamber and wet-stored dialyzer can reduce, but not eliminate the exposure to microbubbles for patients undergoing HD.