Ozone production efficiencies (EN), which can be defined as the netnumber of ozone molecules produced per molecule of NOxoxidised, have been calculated from measurements taken during three intensive field campaigns (one in the spring, EASE 96, and two in the summer, EASE 97 and TIGER 95), at two European coastal sites (Mace Head, Ireland (EASE) and Weybourne, Norfolk (TIGER)) impacted by polluted air masses originating from both the U.K. and continental Europe, as well as relatively clean oceanic air masses from the Arctic and Atlantic. From a detailed wind sector analysis of the EASE 96 and 97 data it is clear that two general types of pollution regime were encountered at Mace Head. The calculated ozone production efficiency in clean oceanic air masses was approximately 65, which contrasted to more polluted air, from the U.K. and the continental European plume, where the efficiency decreased to between 4 and 6. The latter values of ENagree well with literature measurements conducted downwind of various urban centres in the U.S. and Europe, which are summarised in a wide-ranging review table. The EN value calculated for clean oceanic air is effectivelyan upper limit, owing to the relatively rapid deposition of HNO3 tothe ocean. Consideration of the variation of EN with NOx forthe three campaigns suggests that ozone production efficiency is relatively insensitive to both geographical location and season. The measuredEN values are also compared with values derived from steady-state expressions. An observed anti-correlation between EN and measured ozone tendencyis briefly discussed.