This　work　investigates　the　consensus　problem　of　a　linear　multi-agent　system　with　relative-state-dependent　noises　and　time　delays,　where　the　noises　are　proportional　to　the　relative　states　of　agents　and　the　delays　are　time-variant　and　uniform.　The　network　topology　of　the　information　flow　between　agents　is　a　directed　graph.　Based　on　the　analysis　of　stochastic　delayed　differential　equation　and　Lyapunov　stability　theory,　sufficient　conditions　for　the　mean　square　consensus　of　the　multi-agent　system　are　given　by　deriving　out　the　upper　bounds　of　delays　and　the　intensities　of　the　noises.　Serval　numerical　simulation　examples　are　shown　to　verify　the　effectiveness　of　the　theoretical　results.　©　2022　Informa　UK　Limited,　trading　as　Taylor　&　Francis　Group.