We surveyed the impact crater populations of Venus and the Moon, dry targets with and without an atmosphere, to characterize how the 3-dimensional shape of a crater and the appearance of the ejecta blanket varies with impact angle. An empirical estimate of the impact angle below which particular phenomena occur was inferred from the cumulative percentage of impact craters exhibiting different traits. The results of the surveys were mostly consistent with predictions from experimental work. Assuming a sin2 dependence for the cumulative fraction of craters forming below angle , on the Moon, the following transitions occur: <~45 degrees, the ejecta blanket becomes asymmetric; <~25 degrees, a forbidden zone develops in the uprange portion of the ejecta blanket, and the crater rim is depressed in that direction; <~15 degrees, the rim becomes saddle-shaped; <~10 degrees, the rim becomes elongated in the direction of impact and the ejecta forms a "butterfly" pattern. On Venus, the atmosphere causes asymmetries in the ejecta blanket to occur at higher impact angles. The transitions on Venus are: <~55 degrees, the ejecta becomes heavily concentrated downrange; <~40 degrees, a notch in the ejecta that extends to the rim appears, and as impact angle decreases, the notch develops into a larger forbidden zone; <~10 degrees, a fly-wing pattern develops, where material is ejected in the crossrange direction but gets swept downrange. No relationship between location or shape of the central structure and impact angle was observed on either planet. No uprange steepening and no variation in internal slope or crater depth could be associated with impact angle on the Moon. For both planets, as the impact angle decreases from vertical, first the uprange and then the downrange rim decreases in elevation, while the remainder of the rim stays at a constant elevation. For craters on Venus <~15 km in diameter, a variety of crater shapes are observed because meteoroid fragment dispersal is a significant fraction of crater diameter. The longer path length for oblique impacts causes a correlation of clustered impact effects with oblique impact effects. One consequence of this correlation is a shallowing of the crater with decreasing impact angle for small craters.