This dot-probe study assessed anxiety-related biases in visual attentional orienting and spatial motor response selection (motor attention) in high- and low-trait-anxious adults, and whether anxiety-related biases depend on response speed. Emotional-neutral word pairs appeared for 14 or 500 ms, with one word of each pair replaced by a probe. Visual attention bias to emotional words was reflected by faster responses to probes replacing emotional words (spatial correspondence between probe and word positions). Response selection bias was reflected by faster responses when emotional word position (top/bottom screen) spatially corresponded to response position (upper/lower response key). Results revealed anxiety-related bias in visual attention for physical-threat words. In distributional analyses, this bias was associated with slower responses in the 14-ms condition (first task half). Results also revealed anxiety-related effects of spatial correspondence between emotional word and response, which are discussed in terms of increased bias in motor attention towards emotional stimuli in anxiety.