The paper describes a series of analytical and physical model studies conducted to develop an online monitoring system for offshore platforms. An actual offshore jacket platform situated in a water depth of 88m was selected for the study. A detailed three dimensional finite element analysis of the platform involving free and forced vibration revealed that there were dynamic characteristics of the platform which could be used to identify the structural damages in the structure. Having established the feasibility of the method, further work was carried out on a physical model of the platform. Dynamic characteristics of the model were determined by spectral analysis of the response data, simulating various changes including complete and partial damages on individual elements of the model. Simultaneously work was also initiated to apply the results of the physical model tests for interpreting the causes of the changes in the dynamic characteristics using Artificial Neural Networks trained with the database c reated through experiments and analysis. The final outcome of these comprehensive studies was a scheme for integrity monitoring of jacket type of fixed offshore platforms using vibration characteristics of the structure.