In the last decade, event-related potential and functional magnetic resonance imaging studies were very useful in temporal and spatial localization of brain processes involved in the recognition of emotional facial expressions. However, frequency characteristics of the underlying processes have been less studied. Besides, most of the studies did not take into account personality-related individual differences. In this study, effects of explicit and implicit anxiety on the oscillatory dynamics of cortical responses elicited by presentation of angry, neutral, and happy faces were investigated using time-frequency decomposition by means of wavelet transform. Both explicit and implicit anxiety were associated with higher alpha band desynchronization, which was most pronounced during presentation of angry faces. Within theta and delta bands, effects appeared to be opposite for explicit and implicit anxiety measures. In implicitly anxious subjects, frontal delta and theta synchronization upon the presentation of angry and happy (but not neutral) faces was found to be higher than in low anxiety ones, whereas explicit anxiety was associated with a lower theta band synchronization. The results are discussed in terms of conscious and controlled vs. unconscious and intuitive information processing associated with explicit and implicit personality measures.