Understanding of the 'typical' amounts of fog intercepted by different types of cloud forests is hampered by a lack of comparative information on local fog climatology. Usually some kind of 'fog gauge' is used to characterize fog occurrence and amounts. Moreover, wind-driven fog and precipitation are difficult to measure separately and reported measurements of 'fog' often represent a combination of the two. In this paper, the term 'occult precipitation' (HP) is used to represent fog in combination with the horizontal component of wind-driven precipitation (WDR). Collection efficiencies of three widely used types of passive fog gauges, viz. a wire harp (WH) screen, a modified cylindrical gauge (MJU, Juvik-type) and a tunnel gauge (TTG, Daube-type), were derived by comparing the volumes of water collected by the respective gauges with horizontal cloud water fluxes (CWFs) derived from wind speed (u) and the fog liquid water content (LWC) as measured by a cloud particle spectrometer during conditions of fog at a windward cloud forest site in northern Costa Rica. Under conditions of fog-only, the collection efficiencies of the three gauges were linearly related to the horizontal CWF as measured by the gauges themselves. Therefore, additional information on wind speed, droplet size and fog LWC was not needed. During conditions of HP, relative collection efficiencies were derived by comparing the volumes collected by the respective gauge types. The modified Juvik gauge had an efficiency close to 100%, independently of the wind speed and direction, whereas the efficiency of the WH depended critically on the wind speed. The tunnel gauge had an efficiency comparable to that of the Juvik gauge, with some additional catch occurring under conditions of low precipitation angles due to the reclined frontal surface of the gauge. Copyright © 2010 John Wiley & Sons, Ltd.