BACKGROUND: A disturbed circadian rhythm seems to be a causal factor in the occurrence of depressive disorders in patients with Parkinson's disease (PD). The circadian rhythm can be restored with light. Therefore, Bright Light Therapy (BLT) might be a new treatment option for depression in PD patients. METHODS/DESIGN: In this double-blind controlled trial, 84 subjects with idiopathic PD are randomized to either BLT or a control light condition. The BLT condition emits white light with an intensity of 10,000 Lux, while the control device emits dim white light of 200 Lux, which is presumed to be too low to influence the circadian rhythm. Subjects receive 30 min of home treatment twice daily for three months. Timing of treatment is based on the individual chronotype. After finishing treatment, subjects enter a follow-up period of six months. The primary outcome of the study is the severity of depressive symptoms, as measured with the Hamilton Depression Rating Scale. Secondary outcomes are alternative depression measures, objective and subjective sleep measures, and salivary melatonin and cortisol concentrations. For exploratory purposes, we also assess the effects on motor symptoms, global cognitive function, comorbid psychiatric disorders, quality of life and caregiver burden. Data will be analyzed using a linear mixed models analysis. DISCUSSION: Performing a placebo-controlled trial on the effects of BLT in PD patients is challenging, as the appearance of the light may provide clues on the treatment condition. Moreover, fixed treatment times lead to an improved sleep-wake rhythm, which also influences the circadian system. With our study design, we do not compare BLT to placebo treatment, i.e. an ineffective control treatment. Rather, we compare structuring of the sleep-wake cycle in both conditions with additional BLT in the experimental condition, and additional dim light in the control condition. Participants are not informed about the exact details of the two light devices and the expected therapeutic effect, and expectancies are rated prior to the start of treatment. Ideally, the design of a future study on BLT should include two extra treatment arms where BLT and control light are administered at random times. TRIAL REGISTRATION: This trial was registered on ClinicalTrials.gov on May 17th 2012 (ClinicalTrials.gov Identifier: NCT01604876 ).