Exterior greenery systems such as green walls and green roof can provide energy saving and environmental benefits. Several previous studies examined thermal impacts of green roof/walls on a test envelope or simplified buildings in local climates. However, few studies examined the effects of exterior greenery systems considering building types, air conditioning operating conditions, and wall orientations in varied climates. The objective of this study is to investigate cooling energy savings associated with exterior greenery systems for the US Department of Energy (DOE) standard reference buildings, depending on building type, wall orientation, and local climate. Hourly energy consumption was examined for three types of reference buildings: (1) medium office, (2) hospital, and (3) primary school. For the three reference buildings, hourly solar radiation and latent heat transfer rates were calculated for four representative local climates (i.e., hot-dry, hot-humid, cold-humid, and warm-dry). Results show that latent heat transfer due to plant transpiration can dominate heat transfer through the exterior greenery systems. Among the four climates studied herein, heat removal due to evaporation by green walls and roof over the summer months (June to August) is the largest (up to 680 kWh/m2) in Phoenix, while it is in similar range (300–350 kWh/m2) for green walls and roof in other cities. Building type also has notable impacts on the energy performance of the exterior greenery system. Primary school achieves the most significant cooling energy saving in all climates because of high envelope heat gains compared to medium office and hospital. For the primary school in Phoenix, annual cooling energy saving can be maximized up to 27,000 kWh/y with green walls and up to 69,000 kWh/y with a green roof. Latent heat transfer due to evaporation is similar (< 6％) between south and west green walls, implying that utilizing a west green wall could be a good altative or complement to using a south green wall for cooling energy saving.