Planetary　gearbox　has　been　widely　used　in　wind　power,　helicopter,　construction　machinery　and　other　large　complex　equipments.　The　vibration　signals　of　a　planetary　gearbox　are　often　nonlinear　and　nonstationary　when　the　gearbox　works　with　failure.　The　multiscale　entropy　theory　has　been　widely　used　to　measure　the　complexity　of　mechanical　vibration　signals.　Meanwhile,　the　multivariate　multiscale　entropy　(MMSE)　that　evaluates　the　multivariate　complexity　of　synchronous　multi-channel　data　is　introduced　to　the　fault　diagnosis　of　planetary　gearbox　for　using　the　multi-channel　vibration　information　for　improving　the　efficiency　of　fault　diagnosis　as　much　as　possible.　Aiming　at　the　poor　statistical　stability　of　MMSE,　the　multivariate　multiscale　fuzzy　entropy　(MMFE)　was　proposed　and　based　on　this,　a　new　fault　diagnosis　method　for　planetary　gearboxes　was　proposed.　Finally,　the　proposed　method　was　applied　to　the　experimental　data　analysis　of　a　practical　planetary　gearbox　and　the　results　show　its　effectiveness　and　superiority.　©　2019,　Editorial　Office　of　Journal　of　Vibration　and　Shock.　All　right　reserved.