Periodic Leg Movements (PLM) in sleep occur in a wide variety of sleep/wake disorders but their relationship with sleep disturbance, and notably with the concomitant existence of a 'restless legs' syndrome (RLS) remains unclear. We performed 24-h ambulatory polygraphy in a population of 54 consecutive, unselected patients with PLMs (Coleman's index greater than 5/h) who complained of different sleep disorders. A Principal Component Analysis (PCA) was conducted on seven variables from the sample, namely PLM index, patient's age, sleep stage changes per hour, sleep depth index (SWS+PS/TST), diurnal sleep time, number of awakenings exceeding 2 min and presence of a RLS. PCA yielded four independent factors. The PLM index and the changes of sleep stage clustered in a single factor, linking therefore sleep fragmentation to the frequency of PLMs. The second factor appeared to reflect a circadian sleep/wake disorder, combining diurnal sleep time with the number of long night awakenings. The third factor was mainly loaded by the patients' age and the sleep depth index, thus reflecting a well known relationship. Finally, the variable reflecting the existence of a RLS appeared isolated in a single factor, independent from the three previously described. These results confirm and extend the link between PLMs and sleep fragmentation, as well as the recently described dissociation between PLMs and diurnal somnolence. On the other hand, our analysis suggests that in PLM patients the concomitant existence of a RLS is not related to the frequency of occurrence of PLMs, at least when these latter are quantified independently of their arousal index.