Long-term functional outcomes in enthesitis-related arthritis (ERA) is limited from developing countries. We assessed the clinical and genetic factors that predicted the long-term functional outcome in ERA. Patients with ERA having ≥5 years of disease and >16 years of age were included in this cross-sectional study. Data on clinical features within 6 months of disease onset was collected from hospital records. Bath indices, HAQ Disability Index (HAQ-DI) and World Health Organization's Quality of Life (WHO-QOL) were assessed at last visit. Poor functional outcome (PFO) was defined as BASFI > 1.5 or HAQ-DI > 1. Persistent disease activity (PDA) was defined as BASDAI ≥ 4. Endoplasmic reticulum aminopeptidase 1 (ERAP1) and IL-23 receptor single nucleotide polymorphism genotyping was performed with the TaqMan method and HLA-B27 by PCR. One hundred and eighty-one patients [170 male, median (interquartile range) age of disease onset 12.5 (10-15) years, disease duration 7 (5-11) years] were recruited. There was a delay in diagnosis of 3 (1-5) years. The median Ankylosing Spondylitis Disease Activity Score (ASDAS)-ESR, BASDAI, HAQ-DI and BASFI at inclusion were 2.6 (1.8-3.6), 2.6 (1-5.2), 0.5 (0-0.5) and 1.6 (0.3-3.2), respectively. BASFI and HAQ-DI correlated with ASDAS-ESR, ASDAS-CRP and WHO-QOL-BREF. Those with PFO (n = 98) had a longer delay in diagnosis (4 vs 2 years, P < 0.001), lower prevalence of arthritis at onset [odds ratio (OR) = 0.3; 95% CI: 0.1, 0.8], higher prevalence of ERAP1 (rs27044) allele C (OR = 7.2; 95% CI: 1.5, 33.7) and higher disease activity currently. Delay in diagnosis (OR = 1.2; 95% CI: 1.08, 1.4) was the sole predictor of PFO in multivariate analysis. One-third of patients had PDA. Tarsitis at disease onset was the sole predictor of PDA (OR = 2.3; 95% CI: 1.009, 5.4). PFO was seen in one-half of JIA-ERA in the long-term and was associated with active disease with delay in diagnosis as its sole predictor.