26We examined the effects of high-versus low-anxiety conditions during video-based 27 training of anticipation judgements by international-level badminton players facing serves and 28 the transfer to high-anxiety and field-based conditions. Players were assigned to a high-anxiety 29 training (HA), low-anxiety training (LA) or control group (CON) in a pre-training-post-test 30 design. In the pre-and post-test, players anticipated serves either from video under high-and 31 low anxiety conditions or live on-court. In the video-based high-anxiety pre-test, anticipation 32 response accuracy was lower and final fixations shorter when compared to the low-anxiety pre-33 test. In the low-anxiety post-test, HA and LA demonstrated greater accuracy of judgements and 34 longer final fixations compared to pre-test and CON. In the high-anxiety post-test, HA 35 maintained accuracy when compared to the low-anxiety post-test, whereas LA had lower 36 accuracy. In the on-court post-test, the training groups demonstrated greater accuracy of 37 judgements compared to the pre-test and CON. 38 39