In this brief, an online relay autotuning scheme for parameter estimation of a class of time delayed systems is presented. The identification method comprises a relay and a proportional-derivative (PD) controller which helps in yielding sustained oscillation or limit cycles at the system output around the user defined setpoint. Along with the relay and controller parameters, the important limit cycle information are measured and further substituted in the derived set of mathematical expressions required for the estimation of system unknown dynamic in the form of under-damped second-order plus dead-time (SOPDT) model. Simulations of well known model dynamics from the literature are considered to illustrate the efficacy of the proposed scheme in presence of noise and static load disturbance. For the real-time validation of the proposed scheme, a pipe-line water pressure control system is identified as under-damped SOPDT transfer function model. © 2004-2012 IEEE.