It has been established that cortical visual neurons in vivo are in a state of high conductance due to large synaptic input. Nevertheless, little modeling of network dynamics has taken this finding into account. Here we try to begin to fill this gap by extending the mean field theory for a visual cortical hypercolumn presented at SFN 2003 (article at www.arxiv.org/abs/q-bio.NC/0403037) to include conductance-based synapses. Like that work, the theory can describe both population-average properties and neuron-to-neuron variation in responses and correlations. We find contrast-invariant tuning of the average responses, with fluctuations of a size compatible with experiments in the responses of individual neurons. The tuning of the firing irregularity, quantified by the Fano factor F, shows interesting orientation tuning properties, which depend on the synaptic conductance strengths, the post-firing reset level and the synaptic time constant. For experimentally reasonable choices of these model parameters, we...