Public sewage system in rural areas is the basic environmental facility to purify water discharged to local rivers. It is important to forecast water quality using a model to manage the self-purification capacity of local rivers. In this paper, not a modeling but a statistical method of ARIMA time-series was used to forecast water quality of small towns’ sewage. In addition, possible correlations between the five basic water quality assessment parameter (BOD, COD, SS, TN, and TP) were determined. The result of this study give insights to improve management strageties for public sewage systems in rural areas.. This research focused six public sewage systems in rural areas at C-city with water quality assessment data since last five years. According to correlation analysis, it was found there was highly positive relationship between BOD-COD, BOD-SS, COD-SS, COD-TN, COD-TP and TN-TP(the highest of Pearson coefficient r .787 with 99 % confidence level). Based on ARIMA forecasting, the density of BOD and TN would increase and those of COD and SS would decrease.