Commonly used methods for microdialysis recovery measurement are reviewed a
nd the zero flow and no net flux methods are suggested as the most robust i
n practice. Six different mathematical models of microdialysis assumptions
are investigated and compared for varying dialysis probe radius. One transm
itter (dopamine), three metabolites (DOPAC, HVA and 5HIAA) and two drugs (c
affeine and theophylline) were studied. Histology and functional response t
o a drug were measured. Deficiencies were demonstrated for several of the m
odels, the one: best explaining experimental data includes both passive dif
fusion and active tissue regulation in a cylindrical symmetric geometry. Th
e recovery decreased with decreasing probe radius but smaller probes caused
less tissue injury. It Is concluded that a mathematical model of microdial
ysis must include diffusional and physiological processes in order to accur
ately account for experimentally observed phenomena. The experiments also d
emonstrated that, for small brain nuclei, the size of the nucleus may influ
ence the recovery. (C) 2000 Elsevier Science B.V. All rights reserved.