We studied the seasonal population dynamics of Xylosandrus compactus on Coffea canephora for 24 months from June 2013 to May 2015 at the National Coffee Research Institute, Mukono, Uganda. Every month, we randomly sampled 20 infested twigs from different coffee trees on each of three experimental blocks, and dissected them to count different stages of X. compactus per twig. We recorded monthly mean temperature and total rainfall. Seasonal fluctuations in numbers of different life stages of X. compactus inside C. canephora twigs over time were analysed with a general linear model. Ambient temperature and rainfall were used to predict changes in monthly mean counts by means of a multiple linear regression. The population of X. compactus varied significantly across months and its major peak occurred from May to August. This variability was poorly explained by prevailing weather conditions with only the number of dead females correlated negatively with mean monthly temperature. Therefore, it may be suitable to apply management interventions against X. compactus on C. canephora in Uganda prior to May–August of each year to prevent the pest from attaining damaging levels. Further research is necessary to better understand the effect of weather variables on population dynamics of this insect pest.