The sensitivity of source apportionment results to different profiles were tested. Local and non-local mobile source profiles may cause specific apportionment bias. Averaged samples are less affected than daily samples with non-local source profile. US mobile profiles are usable in source apportionment for Chinese samples.