Traditional microdialysis techniques provide qualitative data, although quantitative data are often required for pharmacodynamic analyses. This study evaluated a potentially useful in vivo delivery technique to calibrate microdialysis probes for ethanol. We measured in vivo delivery extraction fractions within subjects across 2 days and found no change over time. We tested the effect of diffusion direction on extraction fraction and found that it was higher for ethanol diffusion out of the probe than for diffusion into the probe, both in vitro and in vivo. The in vivo extraction fraction ratio of diffusion(IN) versus diffusion(OUT) was 0.65+/-0.03. Finally, we predicted extracellular brain ethanol concentrations after 1 g/kg ethanol administration using in vivo delivery, "no net flux" dialysis, or in vivo delivery corrected for diffusion direction with the in vivo extraction fraction ratio. Both in vivo delivery and "no net flux" dialysis predicted brain concentrations that were approximately one-third lower than blood concentrations, whereas the corrected in vivo delivery predicted extracellular concentrations very similar to blood concentrations. We conclude that microdialysis calibration methods for ethanol require a measure of extraction fraction for diffusion into the probe. Further studies are needed to establish whether this effect is common to other alcohols.