Using motion aftereffects (MAE), we quantified the adaptation produced by an expanding concentric ring stimulus. The rings had sinusoidal luminance modulations and were presented in a 5 -diameter aperture against a mean grey background. Two adapting stimuli were used: a concrete stimulus (complete circle) and a phantom stimulus (complete circle minus two horizontal grey 40 -arc sectors either side of the mid-line). After adaptation, subjects indicated if test gratings presented for 0.4s in 30 -arc horizontal sectors were expanding (+) or contracting (-). The subjects adapted for 30s, then indicated the motion direction in 32 tests interleaved with 5s top-up adaptation periods. Adapting temporal and spatial frequencies (TF, SF) were 0-24Hz and 1.5-6cpd. Test TFs were from -1 to 1Hz. Results were obtained for 8 subjects. MAE strength was measured as the test TF perceived as being stationary (i.e. the point of subjective equality (PSE) at which half the tests appear expanding). Concrete and phantom adaptation gave PSEs of 0.1-0.4Hz and -0.1-0.2Hz, respectively. The maximum PSE shifts were at TFs of 8-16Hz (concrete) and 2-4Hz (phantom). Since it is probable that adaptation of local and global motion detectors generates the concrete MAE but only adaptation of global motion detectors generates the phantom MAE, the results suggest that global detectors in the central field are tuned to lower TFs than local detectors. Two further findings were: Stationary control stimuli persistently contracted (PSE 0 to -0.1Hz), possibly due to adaptation during prior locomotion; and (2) phantom adaptation with TFs of 16-24Hz reversed the normal MAE direction, so that slow contraction was perceived as expansion. The different tuning of the concrete and phantom MAEs suggests that local and global motion detectors in the central visual field have different spatiotemporal preferences.