This paper analyzes hourly rainfall data from a collection of tipping bucket rain gauges in British Columbia. The hourly rainfall data are used to define peaks over threshold (POT) rainfall events for durations of 1, 2, 6, 12 and 24 hours. This database is then used to define, on an annual basis, the number of over threshold events, the average magnitude of the over threshold events, and the largest over threshold event. Trend analysis is conducted for these three variables for each duration and for several common analysis periods drawn from the period 1966 to 2005. The identification and estimation of trends is conducted using the Mann-Kendall nonparametric test for trend. The global, or field, significance of the trend results is established using a bootstrap resampling approach. The research reveals generally increasing trends in extreme rainfall, especially for the summer season and for the short duration rainfall events.