Iloprost, a prostacyclin analogue, is a treatment option for surgically unsuitable diabetic chronic critical limb ischemia (CLI), although its outcome is difficult to be anticipated clinically. Whether transcutaneous (tc) oxygen tension (PO2) predicts the response to iloprost in diabetic CLI is unclear at this point and, in that same context, the prognostic role of tc carbon dioxide tension (PCO2), another ischemia-sensitive parameter, is unknown. Supine and dependent tcPO2 and tcPCO2 were measured at baseline and after 4 weeks of iloprost treatment in 31 limbs of 26 type-2 diabetic arteriopaths with CLI not amenable to surgery. Success was defined as pain relief and significant reduction of analgesics. Clinical outcome was stratified by baseline tcPO2 and tcPCO2 tertiles, and likelihood ratios (LR) quantified the increase from pretest chances given a certain result. Iloprost succeeded in 16 (52%) and failed in 15 limbs (48%) and post-treatment tcPO2 followed a parallel course. Failures increased by ascending baseline tcPCO2 and descending tcPO2 tertiles; successes behaved specularly. Predictions of failure based on elevated tcPCO2 (>53 mm Hg) were more efficient than relying on depressed tcPO2 (LR 10.7 vs 3.6); success was almost certain when tcPO2 was >23 mm Hg (LR = 17.8). Dependent determinations were less useful than supine measurements for prognostic use. Elevated tcPCO2 predicted failure efficiently and high tcPO2 was a useful prognostic tool for success of iloprost, suggesting that their combined use may allow better prognostic stratification and improve the therapeutic approach to diabetic CLI.