Vibration-based　structural　health　monitoring　usually　needs　to　extract　vibration　characteristics　from.　operational　vibration　measurements.　The　stochastic　subspace　identification　(SSI)　algorithm　is　an　advanced　technique　for　performing　such　an　operational　modal　analysis.　A　newly　developed　signal　processing　technique,　called　emipirical　mode　decomposition　(EMD),　is　capable　of　dealing　with　non-stationary　signals.　An　EMD-based　stochastic　subspace　identification　procedure　utilizing　operational　vibration　measurements　is　presented　in　this　paper.　The　output-only　measurements　are　first　decomposed　into　modal　response　functions　by　means　of　the　EMD　technique,　on　the　basis　of　specified　intermittency　frequencies.　The　stochastic　subspace　identification　method　is　then　applied　to　the　decomposed　signals　to　yield　the　modal　parameters.　A　case　study　of　the　operational　measurements　from　a　real　bridge　is　presented,　in　order　to　illustrate　the　applicability　of　the　proposed　technique.　It　is　demonstrated　that　the　stable　pole　in　the　stabilization　diagrams　becomes　unique　and　the　vibration　characteristics　are　easily　identified　for　the　decomposed　signals,　bypassing　the　influence　of　other　modal　components　and　fake　frequencies　due　to　unwanted　noise.　(c)　2005　Elsevier　Ltd.　All　rights　reserved.