If neural synchrony provides a meaningful code then one would predict that higher order visual cortical neurons would show robust synchrony when the feedforward signals from lower order neurons are strong and stimulus appropriate. We examined this using a 100 electrode array placed across V1, V2 and DM/V3 (confirmed by histology) in anesthetized, paralyzed bush babies (Otolemur garnetti). Well-isolated units (~50-60 cells/case) were tested with moving gratings differing in orientation, spatial and temporal frequency. Preferred parameters were determined and receptive fields were individually mapped. 1905 cell pairs with overlapping receptive fields were analyzed with joint peristimulus time histograms (JPSTH) (Aertsen, 1989). Neurons were classified as synchronous when the peak duration of the cross-correlation histogram (CCH) was less than 7.0 ms. A neuron was considered to be driving another neuron when the peak duration was more than 7.0 ms. Synchrony was over 19X greater between neurons within a visual area compared to neurons across visual areas and was significantly more frequent between pairs in V1 than between pairs in V2 or DM/V3. The strength of synchrony, reflected by the CCH peak, however, was significantly greater for synchronized pairs in higher then in lower visual areas. There also was a strong correlation between driving efficacy and synchrony, suggesting that the amount of synchrony among neurons in lower areas may determine how effectively spikes are evoked in higher areas. Interestingly, in cases where neuron A appeared to be driving neuron B there often appeared to be two synchronous peaks of similar lag time but opposite sign. We hypothesize that these peaks could reflect feedforward and feedback signals between visual areas. Our results suggest that neuronal synchrony becomes more sparse but stronger in higher visual areas and is strongly correlated with the degree of cooperation between networks within and between these areas.