Suitable for a variety of distributions in the exponential family, generalized linear models (GLMs) use a link function to link a response variable to explanatory factors. For count data, the Poisson distribution is commonly employed; nevertheless, overdispersion results from numerous violations of the distribution's premise that mean equals variance in practical situations. Negative binomial regression is a preferable option in these circumstances. This study uses data from the Selangor Health Department for 2022 and focuses on dengue hotspot Selangor, Malaysia. Using GLMs, it examines meteorological variables such as temperature, precipitation, humidity, and wind speed to find important predictors of dengue occurrence. The variance was more than the mean in the initial Poisson regression, which suggested overdispersion. There was also some moderate collinearity between the climatic variables. Negative binomial regression was thus used, and the results showed that the maximum temperature, minimum temperature, and windspeed significantly impacted dengue cases. This study underscores the need for alternative models like negative binomial regression for over dispersed count data and highlights the critical role of climate factors in managing dengue spread, thus aiding in public health efforts.