Recently, evidence has been raised that long-term changes in the central nervous system are mediated by extrasynaptic spread of neuropeptides ('volume transmission'). To study the effects of volume transmission in the spinal cord we developed the technique of controlled superfusion of the rat cord dorsum. This paper presents quantitative data about the spread, local spinal tissue concentration and redistribution of (2-[125I]iodohistidyl)neurokinin applied for 15, 30 or 60 min to the spinal cord dorsum in concentrations of 0.05 or 50 microM (10 microliters). Analysis of autoradiograms of sagittal and transverse spinal cord sections was done by computer-assisted densitometry. Under all experimental conditions, the spread of radiolabel into the superfused spinal cord segments reached Rexed's laminae V and VI; maximal spread (1.6 +/- 0.3 mm) was measured after superfusion for 30 min. The amount of radiolabel decreased in ventral direction as a function of distance. Highest tissue concentrations of neurokinin A (NKA) were obtained within the superficial spinal cord up to a depth of 0.5 mm and ranged from 700 to 2000 pmol/g following superfusions for 15 or 30 min with 50 microM NKA. Thus, these tissue concentrations were 25-70 times lower than the concentration of NKA in the superfusate. Since pool content was not exchanged, the radioactivity within the spinal cord was lower after superfusion periods of 60 min than after 15 or 30 min. Detection of radiolabel in blood and urine suggests that capillary clearance is relevant and limits the accumulation of the peptide within the spinal cord tissue and the spread into deeper laminae. The controlled superfusion of the rat cord dorsum is a useful method to mimick the spinal release of endogenous neuropeptides such as NKA during intense noxious stimulation, and it can be employed for versatile investigations of the effects of neuroactive molecules on the processing of sensory information in the intact spinal network.