Wind vane motion constants (damping ratio, natural wavelength and decay distance) are derived in a way which can accomodate both mechanical friction and the presence of a propeller. The motion is shown to be insufficiently described by a second order equation because of the way in which the aerodynamic torque changes with angle of attack. This implies that any measurements of vane constants made in the wind tunnel at initial angles of attack above 20° are not representative for the vane. Simple relations between easily measured vane dimensions and motion constants are derived, and vane motion is proved to be independent of the fin area. The WMO requirement for wind vanes is translated into motion constants and shown to be fulfilled for any vane with a damping ratio of 0.30. For turbulence measurements a certain short-wavelength reliability limit for vane-measured spectra is proposed. Experimental comparison of basic fin configurations shows the inferiority of streamlined and splayed fins. General... Abstract Wind vane motion constants (damping ratio, natural wavelength and decay distance) are derived in a way which can accomodate both mechanical friction and the presence of a propeller. The motion is shown to be insufficiently described by a second order equation because of the way in which the aerodynamic torque changes with angle of attack. This implies that any measurements of vane constants made in the wind tunnel at initial angles of attack above 20° are not representative for the vane. Simple relations between easily measured vane dimensions and motion constants are derived, and vane motion is proved to be independent of the fin area. The WMO requirement for wind vanes is translated into motion constants and shown to be fulfilled for any vane with a damping ratio of 0.30. For turbulence measurements a certain short-wavelength reliability limit for vane-measured spectra is proposed. Experimental comparison of basic fin configurations shows the inferiority of streamlined and splayed fins. General...