OBJECTIVES: It has long been suspected that treatment compliance is affected by HRQoL. The aim of this study is to assess how compliance is related to GSS scores and other treatment variables in Glaucoma. METHODS: An observational study was carried out gathering information on the Spanish version of the Glaucoma Symptom Scale (GSS), three compliance scales (Morisky-Green, Haynes-Sackett, & Batalla) and information about disease severity and treatment. Concordance between adherence scales was first assessed. Using the Morisky-Green score as target criteria, a segmentation tree was tested in order to forecast compliance probability. CHAID growing algorithm was used. RESULTS: A sample of 367 patients was recruited with an average age of 68 (±12.1) years from which 57% were females. Most patients (87.7%) were diagnosed of glaucoma, 10.5% were diagnosed of glaucoma and severe PIO and 1.8% presented sever PIO alone. Agreement between adherence scales was poor with a marked superiority of the Morisky-Green questionnaire. The most significant segmentation variable was GSS score, followed by age, glaucoma severity (as assessed by the clinician), number of treatment drugs, and number treatment of drops. Interaction was found between number of drops and other variables in the model. CONCLUSIONS: Using the tree model developed it is possible to predict compliance probability. Self perceived symptom discomfort is found to be the best predictor. Additional evidences of GSS construct validity as a measure of HRQoL have been found.