There exists a great need for automatic precipitation gauges for effective road maintenance during the winter period. These gauges should be inexpensive, not require mains supply, need little attendance, give information about presence of precipitation and determine type (snow, rain or sleet) while there is no need for high accuracy of the precipitation intensity. Light attenuation precipitation sensors (optical gauges) fulfil several of these requirements and are used in the Swedish National Road Administration Road Weather System. The optical gauges measure the time it takes for particles of snow etc. to pass (attenuate) a light beam and relate this time to precipitation type and intensity. The rain precipitation mass is approximately proportional to the accumulated attenuation time. To investigate whether or not optical gauges could also be used for solid precipitation, the precipitation mass for snow, rain and sleet was measured with a reference gauge and compared to the attenuation time. The passage time of individual hydrometeorologic particles (snow, rain and sleet) was compared with precipitation type and wind speed. Air temperature could be used as a rough guide to distinguish three precipitation categories for the following temperatures: rain (>+2 °C), sleet (0 to +2 °C) and snow (≤ 0 °C) At low wind speeds (< 3 m/s) the passage time of individual particles could be used to distinguish between rain and snow. The accumulated attenuation time for the same precipitation mass was approximately 25 and 5 times greater for snow and sleet respectively compared to rain. With the attenuation time for the snow-fall corrected for wind influence the quotient between the attenuation time for snow and rain is decreased from approximately 25 to approximately 12 times.