Nursing personnel in Brazil are usually submitted to fixed 12 h shifts with no consecutive working days or nights. Moonlighting is common in this group, with a consequent increase in the number of working hours. The possibility of sleeping on the job during the night shift in the studied hospitals had already been described. The present study aims to analyze whether the time devoted to daily activities (sleep, rest, leisure, housework, commuting, personal needs, care of children or other people, non-paid work, and study) is related to the number of worked hours and to nap-taking during the night shift. The field study took place at two public hospitals in Rio de Janeiro, Brazil. Workers filled out a structured form on time devoted to the above-mentioned activities for at least four consecutive days. The time devoted to sleep was analyzed according to its occurrence at home or on the job. Workers were classified according to the number of jobs (one job/two jobs) and the time dedicated to work according to the median of the whole series (below the median/above the median). All workers who had at least one working night were analyzed as to nap-taking on the job. They were classified according to the sleep occurrence during the night shift-the sleep group and the non-sleep group, both of which were compared to daytime workers. Statistical treatment of data included non-parametrical procedures. The study group comprised 144 workers (mean age: 35.7+/-10.5 years old; 91% women; 78% nurse assistants, the remainder registered nurses). They recorded their daily activities for 4-11 days; 829 cumulative days were analyzed for the whole group. A total of 165 working nights were analyzed; sleep or rest occurred during 112 (68%) of them, with mean sleep/rest duration of 141+/-86 min. Time devoted to sleep and leisure varied according to the number of working hours, being significantly reduced in those submitted to longer work hours (p < 0.001 and p = 0.002, respectively). Results close to significance point to a reduction in the time dedicated to housework among workers with long work hours (p = 0.053). The time spent on sleep/rest per working night did not differ according to the number of worked hours (p = 0.490). A tendency was observed for those who have two jobs to devote more time to sleep/rest on the job (p = 0.058). The time of personal needs was significantly lower among those who did not sleep on the job as compared to day workers (p = 0.036). The total sleep time was significantly lower among those who did not sleep on the job, as compared to day workers and to those who slept on the job (p = 0.004 and p = 0.05, respectively). As to home sleep length, workers who slept and those who did not sleep on the job were similar and slept significantly less than exclusively daytime workers (p < 0.001 and p = 0.002, respectively). Sleeping on the job during the night shift seems to partially compensate for the shorter sleep at home among night workers and may play a beneficial effect in coping with two jobs.