To calibrate oculokinetic perimetry (OKP) as developed by Damato in terms of conventional perimetric threshold values, 33 eyes with either glaucoma or ocular hypertension were tested with a standard Humphrey Field Analyzer using the Central 30-2 test and twice with a 26-point OKP chart. The frequency of seeing of the OKP test spot was plotted against 30-2 thresholds. This showed a weak relation between the two tests. Subsequently eyes that were considered to have had poor fixation were omitted, but false positive and false negative results still occurred in 19 remaining eyes. To check whether Troxler's effect (local adaptation) might have caused false positives, 6 subjects were tested with OKP, controlling the fixation times. Increasing the fixation time from 'very short' to 2 and 5 seconds yielded more 'not seen' responses in OKP. False negative data were found to occur preferentially in border areas of visual defects. The average frequency of seeing curve over all data showed a 50% frequency of seeing for the OKP stimulus at a 14.6 dB equivalent conventional threshold value. The spreading of the curve was 7 dB (95% confidence interval 28 dB).