The measurement of turbulent fluxes in the atmospheric boundary layer is usually performed using fast anemometers and the Eddy Covariance technique. This method has been applied here and investigated in a complex mountainous terrain. A field campaign has recently been conducted at Alpe Veglia (the Central-Western Italian Alps, 1746 m a.s.l.) where both standard and micrometeorological data were collected. The measured values obtained from an ultrasonic anemometer were analysed using a filtering procedure and three different coordinate rotation procedures: Double (DR), Triple Rotation (TR) and Planar Fit (PF) on moving temporal windows of 30 and 60 min. A quality assessment was performed on the sensible heat and momentum fluxes and the results show that the measured turbulent fluxes at Alpe Veglia were of a medium-high quality level and rarely passed the stationary flow test. A comparison of the three coordinate procedures, using quality assessment and sensible heat flux standard deviations, revealed that DR and TR were comparable, with significant differences, mainly under low-wind conditions. The PF method failed to satisfy the physical requirement for the multiple planarity of the flow, due to the complexity of the mountainous terrain.