Although non-invasive EEG is one of the most widely used tools for studying brain activity in humans, we still lack a clear understanding of how EEG signals are related to the spatio temporal organization of the underlying neuronal activity. In particular, it remains unknown how changes in cortical power and synchrony are reflected in the surface EEG signal. Here, we present an approach which incorporates the power and coherence of local field potentials (LFPs) in the striate cortex (V1) to predict fluctuations in the surface EEG signal. We made simultaneous recordings of neural activity with one surface EEG and multiple intracortical electrodes in two awake monkeys during different behavioral conditions as well as under the effect of local Lidocaine injections. Using a General Linear Model (GLM), we found that both LFP power and coherence conveyed individual information which could be used to accurately reconstruct trial-by-trial fluctuations in EEG power. Furthermore, predictive power of the GLM was very robust across different behavioral conditions but highly dependent on frequency. While EEG power could be modeled with high accuracy in the high frequency regime (R^2=0.719±0.042), predictive power vastly reduced for lower frequencies (R^2=0.0795±0.029). During application of Lidocaine, LFP power was reduced, however, inter-electrode coherence strongly increased, resulting in a scenario where the local cortical area was attenuated, though highly synchronized. Interestingly, the EEG showed an overall increase in power under these conditions, being strongest in higher frequencies, which emphasizes its ability to react to changes in synchrony even when the overall power of cortical activity is diminished. The results of our study emphasize two main p