The role of transformer in electricity network reliability and safe operation is quite crucial. Therefore, their monitoring, maintenance and management is vitally important for utility operators. Simpler, non-invasive and online condition monitoring method which is sensitive to incipient faults is in demand for transformer diagnosis. Although vibration monitoring is considered as a valuable technique in industry for rotating machines, it is least explored for static electrical equipment such as transformer. In this study, core and winding vibrations in a dry type transformer is monitored using an accelerometer to diagnose the winding electrical and mechanical faults. Winding and core vibrations for various degrees of inter-turn fault, axial movement and disc displacement under different load conditions are studied.