Meadow pipit migration and orientation was studied by (1) watching visible migration and measuring flight directions with a compass in NW-Germany (1980-1985) and (2) conducting orientation experiments with caged migrants in radial-perch cages (MERKEL & FROMME, 1958). Free-flying pipits migrated SW in autumn and NE in spring in sunny and overcast weather, but on average there was less migration under overcast. In the field migration peaked in the first two morning hours and was almost over by the end of October. Strong migration always corresponded to a minimal scatter of flight directions. In the orientation cages the birds oriented SW in autumn and NNW in spring, but only in sunny weather during the first three morning hours. Activity decreased markedly later in the day (spring), in November compared to September/October and under overcast compared to sunny skies. The more active the birds were, the better was their orientation with respect to the overall mean direction. In general, meaningful orientation in the cages occurred only when migration intensity in the field was maximal. A lack of well-oriented behaviour in the cage is not necessarily caused by an inability to orient, but may be a consequence of low motivation to migrate. A six hour clockshift of experimental birds did not lead to a clear shift of their directional choices, but produced an axial NE-SW orientation in autumn and disorientation is spring. This result does not support the assumption that diurnal migrants orient primarily by means of a sun compass.