Little information is available on the effects of changes in land use/land cover (LULC) on climate variability in Ethiopia. To characterize this influence, a study was conducted on 30 selected coffee plots along an altitudinal gradient (1500–2100 masl) in Jimma area. The LULC of the transect was characterized using aerial photographs and satellite images and clipped around each coffee plot at the scale of 50, 100 and 200 m radius. To determine the effect of shade (trees), one of the LULC, on microclimate variability, temperature (°C) was recorded both under shade and open area using data loggers along the gradient from June 2012 to 2015. Eight LULC (crop land, pasture land, exotic trees, indigenous trees, river, road, urban and extraction site) were identified in the area. Some of the coffee plots were composed of small areas of trees and large areas of cropland and vice versa. The presence of cropped and pasture land prevented the occurrence of indigenous and exotic trees respectively. Based on LULC, the 30 coffee plots were grouped into three coffee classes (Isolated coffee plots (class1), patch of coffee plots (class2) and coffee plots with contiguous forest (class3). Coffee plots of class3, characterized by high tree density, has lower mean temperature and high relative humidity and wetness duration both during wet and dry season. During the wet season, there was a maximum temperature difference of 1.21 °C among the coffee classes while in the dry season it was 1.03 °C. Furthermore, a mean temperature difference of about 1 °C was observed between open and under shade conditions. Along the gradient, the variation was similar indicating a possibility of developing a shade management strategy as an adaptation option to climate change impact on coffee along an altitudinal gradient.