To improve the scientific basis of distribution network disaster prevention and mitigation decisions, accurate fault statistics parameters of the network are needed. To this end, weather- and time-related fault statistical analysis methods for distribution networks are used. Distribution network fault parameters such as failure and outage rates in the same historical period are calculated according to the different months and weather conditions as well as different line types. The probability density distribution fitting method is proposed for the forced outage time in different weather conditions. The case study of a large distribution network in the southern coastal area of China shows that the monthly distribution of cable faults is not obvious, and the cable faults are less affected by weather. Overhead line faults are greatly affected by thunderstorms and rainstorms, and their distribution is concentrated in May to September, showing obvious single-peak characteristics, and it can be fitted by a Gaussian distribution. The failure and outage rates in thunderstorms and rainstorms is significantly higher than the basic failure and outage rates; and the forced outage time is longer, and can be fitted by Weibull or Gamma distributions. The proposed method has implications for refined fault rate statistical analysis of distribution networks. It can guide the planning and operational risk assessment of distribution networks, and the optimization of fault recovery strategies in adverse weather. This work is supported by the Youth Fund of National Natural Science Foundation of China (No. 51707018).