Background & objectives: Information about the duration of survival of Shigellae in water is speculative. The present study was undertaken to assess the duration of survival of Shigella dysenteriae type 1 and S. flexneri type 2a in the laboratory conditions simulating the aquatic environment, their invasive property and the association of different physico-chemical parameters in the survival process. Methods: Five natural water sources were selected in a diarrhoea prone rural area. Collection of water, determination of physico-chemical parameters and bacteriology were carried out following standard procedures. Filter-sterilised water samples were inoculated with S. dysenteriae type 1 and S. flexneri 2a and survival was monitored by estimating total viable count at regular intervals. Bivariate correlations between the duration of survival and physico-chemical parameters were estimated. Multiple linear regression models were fitted for the duration of survival of the bacteria. Results: All water sources were contaminated with faecal coliforms including Escherichia coli, S. dysenteriae type 1 survived for a mean duration of 3.33 days and S. flexneri 2a for mean 11.167 days in field water samples in laboratory condition. Duration of survival had positive correlation with the initial bacterial counts. In the multiple regression model the strongest predictor of the duration of survival of both S. dysenteriae type 1 and S. flexneri 2a was the concentration of bacteria. Other possible predictors for S. flexneri 2a were Mg and water temperature. Interpretation & conclusion: S. dysenteriae type 1 with epidemic potential survives for shorter duration than S. flexneri 2a. Although some of the physico-chemical parameters had positive relation with duration of survival, the variation of these in natural water samples studied has not caused much variation in the survival in case of S. dysenteriae type 1. In case of S. flexneri 2a, the observed variation in Mg concentration can cause up to 25 days difference in the duration of survival and thus could be a factor determining the endemicity of S. flexneri 2a infection.