Montane cloud forest (MCF) is a rare and vulnerable forest type endangered by anthropogenic habitat loss and climate change impact. Understanding how fog affects the plant community is therefore an urgent task for conservation. We aim to clarify the relationship between fog frequency and environmental factors and evaluate the effect of fog frequency on woody species functional traits. We set up a horizontal transect along the fog frequency gradient in northeastern Taiwan at the elevation around 2000 m asl, in the winter monsoon-affected MCF. In each of 27 plots we collected species composition, leaf and wood functional traits of woody species, and climatic and soil properties. We used Pearson’s correlation to clarify the fog-environment relationships, and simple linear regression to test the relationship between fog frequency and site-specific, inter-specific, and intra-specific community-weighted mean (CWM) of functional traits. We found that in foggier sites, higher precipitation may contribute to the stronger leaching effect, which can lead to lower soil phosphorus content. Also, in this winter-monsoon-influenced cloud forest, foggier sites experience a longer time under subfreezing temperatures (-2.22˚C), explaining the lower specific leaf area (SLA) and higher leaf dry matter content (LDMC) in intra-specific CWM in foggier sites. Also, owing to the unbalanced freezing stress on leaves between evergreen and deciduous species in winter, deciduous species (with high SLA) aggregate more in foggier sites, which may be the reason for higher inter-specific SLA CWM observed there. Although our study is purely observational, we believe that careful selection of plots along the fog frequency gradient in similar elevation allowed us to eliminate most of the confounding environmental factors. It represents unique example of trait changes along fog frequency gradient and brings new ecological explanations. We hope the results can be useful for both theoretical science and also for conservation management decisions.