BACKGROUND: The time distribution of injuries is not random. To assess the potential impact of weather and the phase of the moon on accidents, adjustment for known periodic and nonperiodic factors may be important. We compared the incidence of injuries with quantitative and qualitative weather variables as well as the lunar cycle, after correction for calendar and holiday-related factors. METHODS: We extracted the daily number of trauma patients treated at the emergency department over 36 years (1970-2005) from the trauma database of our regional hospital. For each patient, age, sex, cause of injury, and severity of injury were recorded. This was combined with daily meteorological data including temperature, precipitation, sunshine, humidity, air pressure, and wind as well as the lunar phase. We also related the rate of change of these parameters with the incidence of injuries. A qualitative weather variable derived from temperature, sunshine duration, and precipitation was defined as bad, normal, or good. Periodicities were adjusted for with Poisson regression spline fitting analysis. RESULTS: Several weather variables were related with the number of injuries. For most of these, better weather conditions were associated with an increase in trauma incidence. Good weather, which was present on 16.5% of the days, resulted in 10.1% (9.3-11.4 95% CI) more traumas compared with normal weather. Full moon was associated with a 2.1% (1.1-3.0 95% CI) lower trauma incidence than new moon. CONCLUSIONS: Better weather conditions contribute to an increased incidence of trauma. Full moon is associated with a slightly lower trauma incidence.