The　aim　of　this　study　was　to　quantitatively　investigate　the　effects　of　force　load,　muscle　fatigue,　and　extremely　low-frequency　(ELF)　magnetic　stimulation　on　surface　electromyography　(SEMG)　signal　features　during　side　arm　lateral　raise　task.　SEMG　signals　were　recorded　from　18　healthy　subjects　on　the　anterior　deltoid　using　a　BIOSEMI　Active　Two　system　during　side　lateral　raise　task　(with　the　right　arm　90　degrees　away　from　the　body)　with　three　different　loads　on　the　forearm(0　kg,　1　kg,　and　3　kg;　their　order　was　randomized　between　subjects).　The　arm　maintained　the　loads　until　the　subject　felt　exhausted.　The　first　10　s　recording　for　each　load　was　regarded　as　nonfatigue　status　and　the　last　10　s　before　the　subject　was　exhausted　was　regarded　as　fatigue　status.　The　subject　was　then　given　a　five-minute　resting　between　different　loads.　Two　days　later,　the　same　experiment　was　repeated　on　every　subject,　and　this　time　the　ELF　magnetic　stimulation　was　applied　to　the　subject＇s　deltoid　muscle　during　the　five-minute　rest　period.　Three　commonly　used　SEMG　features,　root　mean　square　(RMS),　median　frequency　(MDF),　and　sample　entropy　(SampEn),　were　analyzed　and　compared　between　different　loads,　nonfatigue/fatigue　status,　and　ELF　stimulation　and　no　stimulation.　Variance　analysis　results　showed　that　the　effect　of　force　load　on　RMS　was　significant　(p　＜　0.001)　but　not　for　MDF　and　SampEn　(both　p　＞　0.05).　In　comparison　with　nonfatigue　status,　for　all　the　different　force　loads　with　and　without　ELF　stimulation,　RMS　was　significantly　larger　at　fatigue　(all　p　＜　0.001)　and　MDF　and　SampEn　were　significantly　smaller　(all　p　＜　0.001).