Abstract: Ma Li, Hu Yue, Xu Yu, Yan Shuxian, Liao Kanghuang Department of Dermatology, Huashan Hospital, Fudan University, Shanghai 200040, China Corresponding author: Ma Li, Email: mary023bb@medmail.com.cn 【Abstract】 Objective To investigate the optimal regimen of narrow-band ultraviolet B （NB-UVB） phototherapy in the treatment of chronic actinic dermatitis （CAD）, and to analyze factors influencing treatment compliance. Methods Demographic data, results of photobiological tests, treatment parameters and clinical responses were collected from CAD patients who received NB-UVB phototherapy in Huashan Hospital affiliated to Fudan University from January 2008 to June 2015, and were reviewed retrospectively. Statistical analysis was done by using two independent samples t-test and chi-square test with SAS9.3 software to compare the clinical data between patients who completed and did not complete the NB-UVB phototherapy. Results A total of 79 CAD patients with Fitzpatrick skin type Ⅳ received NB-UVB phototherapy. Of these patients, 61 （77%） completed the whole treatment, while 18 （23%） dropped out because of intolerance to the NB-UVB radiation. Among the 61 patients who completed the treatment, the average initial, final and cumulative radiation doses of NB-UVB were （0.08 ± 0.01） J/cm2, （0.32 ± 0.08） J/cm2 and （5.9 ± 2.5） J respectively, and patients received （28 ± 8） times of treatment in average. When the radiation dose went up to 0.30 J/cm2, most skin lesions were cleared in 52 （85%） patients. A total of 19 patients received phototesting again after the end of phototherapy. Among 16 patients sensitive to ultraviolet A （UVA） before the treatment, 6 had normal minimal erythema dose to UVA （UVA-MED）, and another 6 had improved UVA-MED after the treatment. Among 16 patients sensitive to UVB before the treatment, 11 got normal UVB-MED and another 3 had improved UVB-MED after the treatment. Univariate analysis showed no significant differences in gender, age, duration of the disease, sensitivity to UVA and UVB radiation, results of photopatch test and patch test between the patients who completed and did not complete the treatment （all P > 0.05）. Conclusions The appropriate NB-UVB phototherapy for CAD patients should start at an initial radiation dose of 0.08 J/cm2 in spring and end at a final radiation dose of 0.30 J/cm2 for about 28 sessions, which can effectively reduce the photosensitivity to both UVA and UVB in CAD patients. Additionally, NB-UVB phototherapy can be applied in CAD patients of different gender, age, disease duration and photosensitive condition.