Paddy production, in the district of Batticaloa, in Sri Lanka, experienced severe challenges due to the rainfall variability in the recent past. The main objective of this paper is to assess the rainfall variability and to identify the relationship between paddy production and rainfall, by means of statistical analysis. Rainfall variability was analyzed using the data of 100 years from 1918 to 2017, of Batticaloa district and the relationship of variables was identified using the data of 40 years from 1978 to 2017. This study adopted a trend analysis using nonparametric and parametric econometric techniques to evaluate the relationship between paddy production and rainfall using EViews 9.0 software. The model was evaluated using Ordinary Least Square (OLS) technique. Results revealed that the annual and seasonal rainfall showing a significant increasing trend. After 1957, rainfall had fluctuated with high variability and rainfall intensity had increased in some years, resulting in flood hazard. On the other hand, dry weather had resulted in long drought condition. Based on the regression results, positive relationship between rainfall and paddy production in Maha season was identified and it is also statistically significant at 1% confidence level. A unit increases in the rainfall in Maha season would increase the paddy production by 0.05 times. Moreover, the rainfall in Yala season negatively correlated with paddy production in Batticaloa. However, it is not statistically significant.