In 12 patients (aged 64 +/- 10 years) with femoropopliteal occlusions (1-27 cm; average, 8.4 cm length) that could not be recanalized by standard guidewire-balloon angioplasty techniques, percutaneous laser-assisted balloon angioplasty was performed by use of a new fluorescence-guided dual-laser system. Plaque detection by 325-nm laser-excited fluorescence spectroscopy provided real-time feedback control to a 480-nm pulsed dye laser (2-microseconds pulses) for atheroma ablation. By means of a common 200-microns optical fiber, after diagnostic fluorescence sensing, computer algorithms directed a fire or no-fire signal (5 Hz) to the treatment laser for selective plaque removal. Laser recanalization (15-50 mJ/pulse) was successful in 10 of 12 patients; this procedure was followed by definitive balloon angioplasty in seven of 12 patients with increased ankle/arm indexes (from 0.60 +/- 0.12 at baseline to 0.84 +/- 0.11 after treatment, p = 0.0043). In laser and balloon angioplasty failures, all femoropopliteal occlusions were heavily calcified, and there were two mechanical guidewire perforations without clinical sequelae. Ablation of calcified lesions required higher pulse energies and greater total energy per centimeter of recanalized tissue (1,837 +/- 1,251 mJ/cm vs. 90 +/- 39 mJ/cm, p = 0.0036). Fluorescence spectroscopy (n = 219 sites) was helpful in flush occlusions and correctly identified plaque, underlying media, and thrombus by changes in fluorescence intensity, shape, and peak position. Thus, when fluorescence-guided laser angioplasty was used in a subgroup of patients refractory to standard angioplasty techniques, primary recanalization and subsequent balloon angioplasty of femoropopliteal occlusions was successful in 83% and 58% of the patients, respectively. Importantly, treatment of heavily calcified lesions accounted for all of the failures and will require modified delivery systems to create larger primary channels and to increase catheter-tip control, which should improve clinical results in the future.