Purpose:　The　purpose　of　this　paper　is　to　explore　the　features　of　health　misinformation　on　social　media　sites　(SMSs).　The　primary　goal　of　the　study　is　to　investigate　the　salient　features　of　health　misinformation　and　to　develop　a　tool　of　features　to　help　users　and　social　media　companies　identify　health　misinformation.　Design/methodology/approach:　Empirical　data　include　1,168　pieces　of　health　information　that　were　collected　from　WeChat,　a　dominant　SMS　in　China,　and　the　obtained　data　were　analyzed　through　a　process　of　open　coding,　axial　coding　and　selective　coding.　Then　chi-square　test　and　analysis　of　variance　(ANOVA)　were　adopted　to　identify　salient　features　of　health　misinformation.　Findings:　The　findings　show　that　the　features　of　health　misinformation　on　SMSs　involve　surface　features,　semantic　features　and　source　features,　and　there　are　significant　differences　in　the　features　of　health　misinformation　between　different　topics.　In　addition,　the　list　of　features　was　developed　to　identify　health　misinformation　on　SMSs.　Practical　implications:　This　study　raises　awareness　of　the　key　features　of　health　misinformation　on　SMSs.　It　develops　a　list　of　features　to　help　users　distinguish　health　misinformation　as　well　as　help　social　media　companies　filter　health　misinformation.　Originality/value:　Theoretically,　this　study　contributes　to　the　academic　discourse　on　health　misinformation　on　SMSs　by　exploring　the　features　of　health　misinformation.　Methodologically,　the　paper　serves　to　enrich　the　literature　around　health　misinformation　and　SMSs　that　have　hitherto　mostly　drawn　data　from　health　websites.　©　2021,　Emerald　Publishing　Limited.