A marked change in ground water quality in the Safford Valley of Graham County, Arizona, averaging approximately +0.129 x 103 mhos electrical conductivity per year and +35 parts per million chloride per year, has been documented between 1940 and 1972 with data from ten long -term sample wells. A chloride change map constructed between these two years shows a general increase of 200 to 400 ppm chloride. The 1972 iso- chemical maps show areas of up to 1600 ppm chloride and 8.0 x 103 mhos electrical conductivity, which is extremely saline and considered threshold level for agricultural waters. The Safford Valley, a structural trough with approximately east -west orientation, averages 12 miles in width and 30 miles in length in the study area. Bounded by typical basin and range province mountains on the northeast and southwest, the valley contains a perennial stream flowing toward the west. A bi- aquifer system constitutes the ground water reservoir of the area with a deep, artesian aquifer of several thousand feet thickness overlaid by a water table aquifer averaging 400 feet in thickness and with the water table rarely over fifty feet from the surface on the eastern end of the valley, deepening to over 5000 feet at the western end. This bedrock -alluvium interface is the lower vertical constraint for the artesian system, thus the thickness of this aquifer increases downstream (to the west). The basin fill consists of a basal conglomerate overlaid by lacustrine evaporite beds, the aquifer cap beds, and recent alluvial material. The artesian aquifer is shown to be up to ten times as saline as the water table aquifer, and appears to increase in temperature and salinity in a downstream direction (corresponding to increasing thicknesses of lacustrine beds included in the extent of this aquifer). The decrement in the water quality of the surficial aquifer seems to be attributable to four major mechanisms. An increase in salinity may be expected from leakage of saline water from the artesian aquifer. Suck leakage would be stimulated by pumping- caused reduction of confining pressure, and by the puncture of the cap beds by deep wells. Water reaching the aquifer from natural recharge may contribute salts to the system. Such recharging water, if passed through soluble beds, could contribute to the salt content of the aquifer. Lateral movement of water through similar deposits may be a contribution, and the concentration and infiltration of agricultural water could also add to aquifer salinity. Ground water applied to the land surface is concentrated by evaporation and dissolves salts in the unsaturated zone as it re- enters the water table aquifer. Iso- salinity and salinity -change maps show the quality situation of the water table aquifer to be broken up into three major sections. From the eastern limit of the study area to Safford, the quality is relatively high and stable. From Safford to Pima there appears a uniform increase of low magnitude but continued decrement. Beyond Pima the area exhibits extremely irregular salinity conditions with marked increases and high salinity gradients. The salinity pattern corresponds to the extent of the underlying artesian aquifer but may be influenced to an unknown extent by the down- gradient transport of salts. The 1972 iso -chemical maps show chevrons of high quality water protruding into the aquifer at points corresponding to the locations of washes. Such wash bottoms are the principal zones of recharge in arid regions. Recharge from the Gila River is of extremely high quality relative to the salinity of the aquifer. There appear no configurations of iso -chemical lines which are attributable to internal movement through saline deposits. The hydraulic gradient of the water table aquifer is relatively constant and follows the gradient of the land surface. Concentration of irrigation water by evaporation and subsequent leaching while in conveyance to the water table seems to increase the salinity of this percolating water by approximately three -fold. The magnitude of this increase at any one point in space and time is a function of the volume of water applied to the land surface, the amount of evaporation, the initial chemical composition of the water, the chemical characteristics of the unsaturated zone through which it penetrates, and the transmission properties of the aquifer. The salinity increase seems significant but the extent of the contribution to the salinity of the aquifer is dependent on the amount of infiltration to the aquifer. This amount is currently undetermined, but is probably a sizable volume -- especially from pre- irrigation applications. A sociologic investigation based on responses from a detailed questionnaire - interview program of 41 farmers (25 percent of the farming population), indicated that there is an awareness of the high salinity of ground water being used for irrigation but relatively little concern about the rate of increase of that salinity. The farmers seem reluctant to leave the area and are willing to take somewhat greater economic losses than expected. Since the farmers of the area are principally Mormon, there may be a tie to this historically Mormon region which is stronger than usual. The economic analysis of the Safford Valley based on the modeling of a "Representative Farm" analog indicates that cotton will remain economical to produce on the basis of the projected salinity trends and ceteris paribus conditions, for a significant time beyond limits of prediction. The analysis indicates that the optimum salt-resistant crops for the area are being cultivated, and that of these, alfalfa, the least tolerant, will cease to be productive in large areas of the valley by 1990. The entire valley will not be able to economically produce alfalfa by 2040, but will remain in production since it is a necessary crop for cotton and the cotton profits should be sufficient to cover the alfalfa losses. Pumping is the only element in the operation of the social, physical and economic systems by which salinity change could be influenced significantly. The area east of Safford is the optimal pumping region while that west of Pima is the worst. The employment of surface water should be maximized, and salt- oriented field methods should be employed. Although agriculture does not seem in danger in predictable time, these practices would increase yield (or slow the decrease) and postpone the day when farming will no longer by profitable in the Safford Valley of Graham County, Arizona.