To　ensure　smoothness　of　synthesized　sign　language　sentence　in　video　format,　which　is　concatenated　with　sign　language　video　clips,　this　paper　proposes　a　kind　of　movement　epenthesis　approach　based　on　prosody　controlling.　The　proposed　method　decides　number　of　transition　frames　according　to　sign＇s　velocity　of　pre-video　clip　and　post-video　clip,　computes　appropriate　locations　on　the　motion　trajectories,　synchronizes　behavior　of　different　models　while　the　transition　units　are　integrated.　Experiment　results　show　that　the　synthesized　sign　language　sentence　is　smooth　and　natural　as　expected.　©　2011　IADIS.