To study the economic benefits of controlling salinity in residential water supplies, we surveyed households and appliance repair shops along the Arkansas River Basin in Colorado, where salinity ranges from 100 to 3600 mg/L. To avoid a downward bias on estimated appliance lives, we obtained and used data on both ages of in-service appliances and ages at failure of failed appliances. We adapted the accelerated testing method to model the effect of salinity on appliance lives. Dishwashers, water heaters, garbage disposers, water softeners, and evaporative coolers showed statistically significant reductions in service life with increasing salinity. In comparison with the most cited previous study, we found no statistically significant effects for some appliances; for appliances common to both studies our estimates of salinity damages are one third or less as high. These differences may originate from inclusion of in-service appliances or reduced damage due to technological improvements. ¿ 2000 American Geophysical Union