OBJECTIVE To evaluate the clinical impact of computer-assisted fluorescein perfusography in peripheral arterial occlusive disease (PAOD). METHODS Foot and calf skin perfusion was visualised by intravenous fluorescein injection. Fluorescein influx was recorded photographically and converted into functional images of fluorescein appearance times (AT) by means of digital film processing. METHODS Vascular Laboratory of Clinic for Vascular Disease. METHODS 249 patients with PAOD. Among 481 limbs studied, 83 legs presented with patent arteries, 70 with asymptomatic obstructions (Stage I), 170 with claudication (Stage II) and 158 with rest pain and skin lesions (Stage III/IV). METHODS Forefoot and calf mean ATs and standard deviations (SD) served as arbitrary measures of regional skin perfusion rates and their homogeneity, respectively. RESULTS In the control legs, a homogeneous and fast fluorescence appearance was observed (medians at the foot: AT33.4 s, SD 3.6). In stage II disease, AT (39.9 s, SD 5.6) were slightly impaired as compared to limbs with patent arteries or stage I disease (p < 0.01). Ninety-seven out of the 158 legs in stage III/IV could be managed by conservative therapy. According to fluorescein-perfusography, they did not differ from stage II disease (AT 38.8 s, SD 6.1). Sixty-one limbs were clinically affected by critical ischaemia. They exhibited a markedly delayed and heterogenous fluorescein influx at the foot (AT 77.3 s, SD 26.5, p < 0.01 vs all other groups). Non-fluorescent areas occurred in 53% compared to only 1% of limbs with and without critical ischaemia, respectively. Retrospectively, predictive values of fluorescein perfusography in identifying a critical limb ischaemia (accuracy 93%) were superior to the ankle systolic arterial pressure determination (accuracy 80%). CONCLUSIONS Fluorescein perfusography seems to be of diagnostic and prognostic use in PAOD in stage III/IV where inflammatory and ischaemic patterns of dye appearance can be distinguished.