Abstract Correction of the I-V curve permits the comparison of curves measured under different conditions for photovoltaic (PV) panels' health monitoring purpose. IEC 60891 has defined three standard procedures named 1, 2 and 3 for the correction. They were initially designed to correct the curves of healthy PV panels. However, their performance, when applied to I-V curves measured on faulty panels, is rarely discussed. This work evaluates these correction methods on I-V curves simulated under different environmental conditions and for five types of defects of varying severity. The results show that procedure 3 has a relatively better overall performance but is not suitable for rapid application in the field as it requires the determination of reference curves. It is found that procedures 1 and 2 could introduce distortion of the curve's shape, with a relative error of up to 13.8% and 6.4%, respectively. A misestimation of 9.1% for key parameters of the curve has been observed, when using procedure 2 for maximum power. Based on the performance analysis, a new correction method is proposed to fit the corrected voltage. It can reduce the curve's average correction error by 31.3% compared to the original single curve correction method. Challenges and directions for future work are also presented.