Background: Elder abuse is an important public health problem hidden from public view in spite of its detrimental effect on physical and mental health. In the last three decades, there is responsiveness toward abuse of children and spouses worldwide but even today abuse and neglect of aged people seem to be the matter of least concern. This study aims to reveal the magnitude of elder abuse and its predictors., Methods: This is a cross-sectional study carried out in a community setting among elderly people (60 years and older) from October 2015 to September 2017 in a rural area of West Bengal, eastern part of India. Elder abuse was elicited by questions adapted from the Actual Abuse Tool. Data were analysed using SPSS (version 16.0). Logistic regression analysis was performed to reveal the factors associated with elder abuse., Results: The study found that 25.6% of the respondents had experienced abuse, psychological abuse being the commonest type. Disability was the most common perceived risk for abuse. Univariate logistic regression showed higher age (70 years and older), female gender, who were Scheduled caste, living in a nuclear family, having more than three children, financially and functionally dependent, who was suffering from depression and multimorbidity had increased odds of being abused. In multivariable logistic regression, people aged above 70 years, living in a nuclear family and dependent for instrumental activities of daily living remained significant predictors of abuse., Conclusion: Around one in four of the respondents was abused in a 1 year duration which implies the urgent need for action. Routine screening for elder abuse at ground level and abuse prevention campaigns can be an important measure to alleviate this hidden problem. Also, there is a need to explore culturally specific forms of elder abuse to provide an age-friendly environment., (© 2020 Japanese Psychogeriatric Society.)