Volumetric soil water content, soil water potential (SWP), leaf water potential (LW?) andrnchlorophyll fluorescence were measured to evaluate physiological responses and adaptability tornwater deflcits in the wild arabica coffee populations along climatic gradients in the southeasternrn(SE), South and Southwestern (SU] of Ethiopia. The sites selected were Harena (Bale) forestrnji-om the SE, Bonga and Shekoforestsji-om the South and Yayuforestji-om the SWregions. Eachrnof the four sites was sub divided in to three sub sites that represent different coffie accessionsrn(sub populations or land races). Therefore, a total of 12 coffee accessions or sub populationsrnwere studied in situ.rnThe soil moisture at two distances ji'Oln the individual coffee trees, wet and dly soil moisturerncontrasts as well as diurnal variations were measured and analyzed. At the same time, soilrnmoisture deeper than the slilface layer was also measured. The measurement of soil waterrnpotential took place at three depths during the wet season but at one depth in the dry season. Inrnthe case of leaf water potential, both predawn and midday as well as diurnal changes werernrecorded in the d,y season but only midday values were recorded in the wet season.rnDiurnal chlorophyll fluorescence yield and light responses were measured for light adaptedrnleaves. Comparisons were made between leaves of three age groups (older, medium aged andrnyounger), wet and dry seasons as well as shade and sun exposed coffee trees.rnThe wet season in Bale differs ji-om that of the three sites (Bonga, Sheko and Yayu). It was dry inrnBale during the two periods of measilrements. Therefore, wet-dlY comparison does not includernBale. Almost similar soil moisture content and soil water potential values were found in Bonga,rnSheko and Yayu during the wet season. But considerable variation was found in most of thernsubsites during the dly season and maximum and minimum values were recorded in Bonga 2 andrnSheko 3 respectively. When the soil moisture values of all the sites were compared based on thernmean values of their sub sites, Bonga and Yayu had the highest and lowest soi/moisture contentrnill the d,y season, respectively. Minimulll and maximum range of wet and d,y seasons' soilrn1rnmoisture content was also found in Bonga and Yayu respectively. Bonga had the highestrnpredawn leaf water potential (LWP) followed by Yayu. The highest and the lowest Midday LWPrnvalues were recorded in Bonga and Sheko respectively. The highest and the lowest diurnalrnranges of LWP were recorded in Yayu and Bale respectively. Based on their LWP values, Yayurnand Sheko coffee populations seem to be drought avoiders (by higher re-saturation capacity)rnwhereas, Bale coffees seem to be drought tolerant. Most probably, in Bonga (the wettest site),rnthe coffee populations were not too stressed to display some drought resistance mechanisms.rnSignificant difference was found beflVeen the wet and d,y seasons' midday LWP values.rnChlorophyll fluorescence yield of coffee leaves of different age groups did not show significantrndifference (P=0.05). But Chlorophyll fluorescence yields of most of the sites and sub sites werernsignificantly different (P=0.05). Higher value of fluorescence yield was found in the d,y seasonrnthan the wet season. Shade coffees had higher values of fluorescence yield than sun exposedrncoffees. The fluorescence yield was linearly and weakly correlated to both soil moisture andrnLWP but the correlation to leaf water potential was relatively higher. A linear, somewhat strongrncorrelation was found between soil moisture and leaf water potential. The strongest correlationrnwas found beflVeell soil moisture content and soil water potential. The grand mean soil moisturernof all the coffee forests for the wet and d,y seasons were 47.91% and 17%, respectively.rnConcerning the sub sites, the highest level of stress wasfollnd in Sheko 3and the lowest in Bongarn2, whereas among the sites, the highest level of stress was found in Sheko and Yayu and thernlowest ill BOllga for the d,y season.