Key points: Highlanders develop unique adaptative mechanisms to chronic hypoxic exposure, including substantial haemoglobin and haematocrit increases. However, a significant proportion of populations living permanently at high altitude develop maladaptive features known as chronic mountain sickness (CMS). This study aimed to assess the effects of permanent life at high altitude on clinical and haemorheological parameters (blood viscosity and red blood cell aggregation) and to compare clinical and haemorheological parameters of dwellers from the highest city in the world according to CMS severity. Blood viscosity increased with altitude, together with haemoglobin concentration and haematocrit. At 5100 m, highlanders with moderate-to-severe CMS had higher blood viscosity mainly at high shear rate and even at corrected haematocrit (40%), with a lower red blood cell aggregation. Blood viscosity may contribute to CMS symptomatology but the increased blood viscosity in CMS patients cannot solely be explained by the rise in haematocrit. Abstract: Chronic mountain sickness (CMS) is a condition characterised by excessive erythrocytosis (EE). While EE is thought to increase blood viscosity and subsequently to trigger CMS symptoms, the exact relationship between blood viscosity and CMS symptoms remains incompletely understood. We assessed the effect of living at high altitude on haemoglobin, haematocrit and haemorheological parameters (blood viscosity and red blood cell aggregation), and investigated their relationship with CMS in highlanders living in the highest city in the world (La Rinconada, Peru, 5100 m). Ninety-three men participated in this study: 10 Caucasian lowlanders, 13 Andean highlanders living at 3800 m and 70 Andean highlanders living at 5100 m (35 asymptomatic, CMS score ≤5; 15 with mild CMS, CMS score between 6 and 10; 20 with moderate-to-severe CMS, CMS score '10). Blood viscosity was measured at native and corrected haematocrit (40%). Haemoglobin concentration and haematocrit increased with the altitude of residency. Blood viscosity also increased with altitude (at 45 s−1: 6.7 ± 0.9 mPa s at sea level, 14.0 ± 2.0 mPa s at 3800 m and 27.1 ± 8.8 mPa s at 5100 m; P ' 0.001). At 5100 m, blood viscosity at corrected haematocrit was higher in highlanders with moderate-to-severe CMS (at 45 s−1: 18.9 ± 10.7 mPa s) than in highlanders without CMS (10.2 ± 5.9 mPa s) or with mild CMS (12.1 ± 6.1 mPa s) (P ' 0.05). In conclusion, blood viscosity may contribute to CMS symptomatology but the increased blood viscosity in CMS patients cannot solely be explained by the rise in haematocrit.