BACKGROUND:The γ process in core-collapse supernova explosions is thought to explain the origin of proton-rich isotopes between Se and Hg, the so-called p nuclei. The majority of the reaction rates for γ-process reaction network studies has to be predicted in Hauser-Feshbach statistical model calculations. Recent investigations showed problems in the prediction of α widths at astrophysical energies. This impacts the reliability of abundance predictions in the upper mass range of the p nuclei.PURPOSE: Measurement of the 127 I(α,γ) 131 I and 127 I(α,n) 130 I reaction cross sections close to the astrophysically relevant energy range to test the predictions, to derive an improved reaction rate, and to extend the database required to define an improved global optical α+nucleus potential.METHODS: The cross sections were derived using the activation technique, the yield of the emitted γ and characteristic X-ray photons were measured using a LEPS and a HPGe detector.RESULTS: The cross sections of the 127 I(α,γ) 131 Cs reaction have been determined for the first time, at energies 9.50 ≤ Ec.m. ≤ 15.15 MeV. The 127 I(α,n) 130 Cs reaction was studied in the range 9.62 ≤ Ec.m. ≤ 15.15 MeV. Furthermore, the relative intensity of the 536.1 keV γ transition was measured precisely, its uncertainty was reduced from 13% to 4%. The results were then compared to Hauser-Feshbach calculations which were also used to extend the cross sections into the astrophysically relevant region and to compute the reaction rate.CONCLUSIONS: The comparison to statistical Hauser-Feshbach model calculations showed that the α width can be described well in the measured energy range using a standard, energyindependent global optical potential. The newly derived stellar reaction rates at γ-process temperatures for 127 I(α,γ) 131 I and its reverse reactions, nevertheless, are faster by factors 4 − 10 than those from previous calculations, due to further improvements in the reaction model. The importance of the inclusion of complete level schemes into the Hauser-Feshbach calculations is illustrated by comparing the impact of two different level schemes, one of them extending to higher excitation energy but not containing all relevant levels.