This paper analyzes the effect of digital differentiation on the measurement of flicker emission from grid-connected wind turbines (WTs). The flicker measurement procedure is defined in the IEC 61400-21 standard, which includes the estimation of a fictitious grid for the measurement of voltage fluctuations exclusively generated by a WT. The study focuses on two aspects of the implementation of the derivative of the line current that affect the estimation of the fictitious grid, namely the bandwidth and the phase response of the digital differentiator. The experiments were performed under two different frameworks, by using simulated simple waveforms and voltage and current recordings from an actual WT. The work cautions about the adverse effect of the phase response of the even-length finite impulse-response differentiators. It also recommends implementing digital differentiators with an accurate bandwidth to derive potential spectral components in the current in the whole bandwidth required by the procedure. This aim can be achieved by increasing either the length of the filter or the sampling rate of the processing. The conclusions should help in accurately implementing the procedure and reducing the deviations found in the flicker emission measurements for grid-connected WTs.