We present initial results from a statistical study of Cluster multispacecraft flux transfer event (FTE) observations at the high-latitude magnetopause and low-latitude flanks from February 2001 to June 2003. Cluster FTEs are observed at both the high-latitude magnetopause and low-latitude flanks for both southward and northward IMF. Among the 1222 FTEs, 36%, 20%, 14%, and 30% are seen by one, two, three, and four Cluster satellites, respectively. There are 73% (27%) of the FTEs observed outside (inside) the magnetopause, which might be caused by the motion of FTEs toward the magnetosheath when they propagate from subsolar magnetopause to the midlatitude and high-latitude magnetopause and low-latitude flanks. We obtain an average FTE separation time of 7.09 min, which is at the lower end of the previous results. The mean BN peak-peak magnitude of Cluster FTEs is significantly larger than that from low-latitude FTE studies. FTE BN peak-peak magnitude clearly increases with increasing absolute magnetic latitude (MLAT), it has a weaker dependence on magnetic local time (MLT) with a peak near the magnetic local noon, and it has a complex dependence on Earth dipole tilt with a peak at around zero. FTE periodic behavior is found to be controlled by MLT, with a general increase of FTE separation time with increasing MLT, and by Earth dipole tilt, with a peak FTE separation time at around zero Earth dipole tilt. There is no clear dependence of FTE separation time on MLAT. There is a weak increase of FTE BN peak-peak magnitude with increasing FTE separation time, and we see no clear dependence of it on FTE BN peak-peak time. When no FTE identification thresholds are used, more accurate calculations of some FTE statistical parameters, including the mean BN peak-peak time, can be obtained. Further, comparing results with different thresholds can help obtain useful information about FTEs.