Major Depressive Disorder (MDD) is a broad heterogeneous construct resolving into several symptom-clusters by factor analysis. The aim was to find the factor structures of MDD as per Montgomery and Asberg Depression Rating Scale (MADRS) and whether they predict escitalopram response. In a longitudinal study at a tertiary institute in north India, 116 adult out-patients with non-psychotic unipolar MDD were assessed with MADRS before and after treatment with escitalopram (10-20mg) over 6-8 weeks for drug response. For total 116 patients pre-treatment four factor structures of MADRS extracted by principal component analysis with varimax rotation altogether explained a variance of 57%: first factor 'detachment' (concentration difficulty, lassitude, inability to feel); second factor 'psychic anxiety' (suicidal thoughts and inner tension); third 'mood-pessimism' (apparent sadness, reported sadness, pessimistic thoughts) and fourth 'vegetative' (decreased sleep, appetite). Eighty patients (68.9%) who completed the study had mean age 35.37±10.9 yrs, majority were male (57.5%), with mean pre-treatment MADRS score 28.77±5.18 and majority (65%) having moderate severity (MADRS <30). Among them 56 (70%) responded to escitalopram. At the end of the treatment there were significant changes in all the 4 factor structures (p<0.01). Vegetative function was an important predictor of response (p<0.01, odd's ratio: 1.3 [1.1-1.6] 95% CI). Melancholia significantly predicted non-response (p=0.04). Non-psychotic unipolar major depression having moderate severity in north Indian patients as per MADRS resolved into four factor-structures all significantly improved with adequate escitalopram treatment. Understanding the factor structure is important as they can be important predictor of escitalopram response.