To better assess potential earthquake hazards requires a better understanding of fault friction and rupture dynamics. Critical slip‐weakening distance () as one of the key friction parameters, however, is hard to determine on natural faults. For strike‐slip earthquakes, we may directly estimate the from —the double near‐fault ground displacement at the time of the peak velocity (Fukuyama and Mikumo, 2007). Yet near‐fault observations are very few, and, thus, there were only limited earthquakes with such estimation. In 2014, an 6.2 strike‐slip event—the Ludian earthquake—occurred in southwest China. The strong‐motion station (LLT) that is from the fault recorded the earthquake and enabled us to estimate from the accelerograms. We inspect the polarity of the accelerometers and compare the integrated velocities with waveforms of nearby broadband stations. We also analyze the particle motion at the LLT station and retrieve the earthquake initiation at the intersection of the conjugated faults. We then apply the baseline correction to the seismograms, recover the ground velocities and displacements, and obtain the value of at the station. The recovered final displacements are compared with the predicted ground displacements of a finite‐fault model. The discrepancy of fault‐parallel displacements might imply limited underestimation of , and the estimated upper limit is 0.3 m. Comparison between the and final slip on the fault patch follows the scaling of previous larger earthquakes. Analysis of the near‐fault accelerometer data enhances our understanding on the earthquake source of the Ludian earthquake. This case extends the lower magnitude boundary of the values obtained from natural faults and opens a window into the friction property in the seismically active region.