Urban　rapid　railway　system　is　a　major　form　of　transportation　for　a　metropolis.　Effective　use　of　urban　rapid　railway　systems　requires　effective　connections　with　other　transportation　modes.　Different　attributes　between　arrival　and　departure　access　modes　are　analyzed　through　a　trip　survey　in　Beijing.　It　is　found　that　the　access　mode　and　departure　mode　show　significant　differences　in　terms　of　distance　and　time.　Using　the　data　fusion　method,　a　nested　Logit　model　of　access　mode　selection　is　formulated,　and　the　influences　of　three　factors,　access　time,　access　cost　and　access　distance　to　the　access　mode　selection,　are　analyzed.　It　is　shown　that　these　three　factors　are　all　negatively　correlated　with　the　access　mode　selection.　The　access　mode　selection　appears　more　sensitive　to　the　fee　for　bus　access.