The aim of our study was to identify the factors substantially affecting day-to-day variability in O 3 concentrations in Czech mountain forests and to describe their influence in detailed, quantitative way. We examined the effects of meteorology and ambient NO x recorded in regular long-term continuous monitoring at five mountain forest sites representing different regions, covering both polluted and relatively unpolluted areas over the time period of 1992-2018. To investigate the association between ambient O 3 concentrations on one hand, and precursor NO x concentrations, and meteorology on the other hand, we used a generalized additive model, GAM, with semiparametric (penalized-spline-based) components to capture properly the possible departures from linearity that is not captured by traditional linear regression approaches. Our results revealed that the O 3 concentrations showed significant associations with all selected explanatory variables, i.e., air temperature, global solar radiation (GLRD), relative humidity, and NO x . Apparently, both meteorology and air pollution are highly important for day-to-day O 3 concentrations, and this finding is consistent for all five rural sites, representing middle-elevated forested mountain areas in Central Europe. In addition to individual variables, we were able to detect interactions between three pairs of explanatory variables, namely temperature * GLRD, temperature * relative humidity, and GLRD * relative humidity. Moreover, we confirmed non-linear O 3 behavior toward all individual explanatory variables.