Atomic diffusion – important in many contexts in stellar astrophysics and an important thread running through this meeting – is most spectacularly observable in the atmospheres of some A stars. The magnetic Ap stars and the non-magnetic Am stars show directly abundance anomalies caused by gravitational settling and radiative levitation. Over the last decade spectroscopic studies have begun to provide maps of abundance distributions in the magnetic Ap stars in three dimensions. Interestingly, high radial overtone p-mode pulsations in roAp stars have also given three-dimensional views of the stellar atmospheres with studies of rotational and line profile variations of pulsation amplitudes and phases. These detailed looks at the effects of microscopic atmospheric changes in the strongly non-LTE and magnetic upper atmospheric layers of Ap stars provide perhaps the most exciting challenge to atomic diffusion theory in terms of detailed explanation and prediction. Am stars were at one time thought not to pulsate because of gravitational settling of He from the He ii ionization zone that provides the κ-mechanism driving for δ Sct pulsations in A stars. In the last few years we have found with SuperWASP and Kepler observations that many Am stars do pulsate. More than half of all A stars pulsate at Kepler micromagnitude precision, yet there is a subset of A stars that truly do not pulsate at that level. Are these Am stars with the strongest signature of atomic diffusion? Is atomic diffusion the reason for the pulsational stability of these stars? The answers are not yet known.