Thirty-seven nurses with shift work, working under a rotating three-shift system, and 37 nurses without shift work, having worked during the daytime for one or more years prior to this study, were examined to assess the effects of shift work on cardiac autonomic and neuromotor functions. Their ages ranged from 25 to 58 yr. The electrocardiographic (ECG) R-R interval variability, %LF and %HF (i.e., proportions of sympathetic and parasympathetic activities, respectively), and LF/HF ratio were computed by means of autoregressive spectral and component analyses. The %LF and LF/HF ratio were significantly larger in the nurses with shift work than in those without shift work, although there was no significant difference in the heart rate-corrected QT interval on ECG between them. And, hand-ear coordination differed significantly between the nurses with shift work and those without. Despite the presence of potential selection bias, it is suggested that shift work in nurses may cause not only a long-term sympathodominant state but also some neuromotor impairment.