Objective: How the brain achieves integration of temporally dispersed information is one of the enigmas in the neurosciences. By combining feature fusion with transcranial magnetic stimulation (TMS), we show that individual visual features are stored for several hundred milliseconds before being condensed into a conscious percept. Methods: We presented two verniers with opposite offset directions in rapid succession at the same retinotopic location. Because of the short durations (30ms each) only one fused vernier is perceived. The perceived offset of the fused vernier is a combination of the offsets of the two verniers which themselves are consciously not accessible. We then applied TMS at different times after the onset of the first vernier over the occipital or frontal cortex. Six healthy paid subjects contributed with 120 trials per condition each to the study. Results: When TMS was applied occipitally from 45ms to 95ms after the onset of the first vernier, the second vernier dominated significantly (p<.05). For later onset asynchronies 170ms up to 320ms, the first vernier dominated significantly (p<.05). Applying TMS frontally no such modulation can be found. Conclusion: Even though the individual verniers are not consciously accessible, they still can be manipulated by TMS for a period of more than 300ms. Hence, our results demonstrate long lasting feature memory and an extended integration process of unconscious feature traces in that is based on early visual cortex rather than frontal areas. Acknowledgement: This work was supported by the SNSF (Swiss National Science Foundation).