Earlier we reported that the urinary excretion of 6-sulfatoxymelatonin (aMT6s) displayed seasonal rhythms in laboratory rats and hypothesized that the horizontal intensity H of the geomagnetic field may act as seasonal zeitgeber. To test this, long-term experiments were performed with female Sprague-Dawley rats. In experiment I (n=12: 1997-1999) nocturnal aMT6s displayed a winter-summer increase by 30% and a rhythm with a phase-length of 24 months peaking in July 1998. In experiment II (n=12; 1999-2000) the winter-summer increase amounted to 40%. The estimated rhythm had a phase-length of 18 months with a peak in September 2000. Compared to experiment I both the rhythm-adjusted mean (MESOR, + 28%) and amplitude (+68%) were elevated. In experiment III (n=30; 2003-2004) the winter-summer increment was just 20%. A circannual rhythm with a peak in April/May was found. The MESOR was 13% higher than in experiment I but the amplitude was depleted (– 14%). In experiment IV (n=15; 2005-2006) a slight winter-summer increase (+15%) was found and a low-amplitude rhythm of 24 months phase-length peaking in June 2006. The MESOR was similar to experiment I but the amplitude was depressed (– 36%).These results demonstrate that female rats within two years of age show elevated aMT6s during summer/spring which supports our initial hypothesis. The apparent inter-experimental amplitude variation indicates the involvement of additional variables. Based on our initial hypothesis, we postulate an involvement of the solar cycle affecting H leading to year to year variations and present supportive analyses.