Under the background of global climate change, the frequency of extreme rainstorm events increases. It is of great significance to study the formation mechanism of flood peak in small watershed for rainstorm flood prediction, disaster prevention and mitigation. Based on the extraordinary rainstorm event caused by Typhoon "Lekima" in Linqu County, Weifang City, Shandong Province in 2019, 17 small watersheds near the rainstorm center were selected, and the response characteristics of the flood peak intensity and sensitivity of the small watersheds to the vegetation and terraces were studied by the methods of field investigation, remote sensing interpretation and correlation analysis. The results showed that in this rainstorm event, the peak discharge of small watershed was 2.36~56.50 m³/s, the peak modulus was 8.00~48.89 m³/(s·km²), and the flood index was between 3.61 and 4.55. Under the condition of similar rainfall, the flood peak modulus, flood index K had significant positive correlation with proportion of sloping farmland (p<0.05), and the flood peak modulus had a significant negative correlation with the proportion of forest grassland and terrace (p<0.05), the flood index K had negative correlation with the proportion of forest grassland and terrace. With the increasing of individual rainfall, the influence of vegetation and terraces on flood peak in small watershed presented a nonlinear feature of increasing first and then decreasing. Under the condition of extreme rainstorm, the sensitivity of surveyed watershed to flood was "Average". The ability of small watershed to cope with extreme rainstorm could be improved by increasing the area of forest, grassland, terrace and other land types.