Abstract A control chart for non-normal correlated data under repetitive sampling is presented in this manuscript. A Burr distribution is employed to model the non-normal distribution. The proposed control chart is based on two sets of control limits whose coefficients are determined by considering the in-control average run length. The tables of the out-of-control average lengths for various shifts, sample size, and correlations are presented in the paper. The proposed control chart is found to be more efficient than the existing control chart based on single sampling in detection of small process shifts.