A variety of statistical procedures was used to analyze data which measure water quality in the Connecticut River. The procedures included analysis of variance, regression and stepwise regression analyses and analyses of time series. The following observations may be made on the basis of the conclusions of the study: (1) Dissolved oxygen levels vary significantly among days and among sampling stations along the Connecticut River, (2) there was no significant variation in dissolved oxygen levels at either of the two sampling depths within a sampling station although there was a significant variation between the two depths, and (3) several different time series models and their requisite computer programs were developed in order to determine which model best fitted the six years of weekly coliform data available. From the time series analysis the coliform content is shown to be increasing from year to year. Finally, a model was developed using regression analysis to predict fecal coliform content from total coliform content. (Author)