Langham Place is a typical case of intensive pattern of rail transit complex in Hong Kong.Based on the research and analysis on pedestrian distribution,this paper found that spatial factors,such as spatial integration,relationship with building entrances,distance to escalators,level,and attraction of function,had high correlation with pedestrian distribution.The multivariate linear regression analysis showed that pedestrian distribution could be explained well and the coefficient of determination could achieve at 61%,when considering the above spatial factors.The research adopted "gate count" and "trace observation" to record the pedestrian flow and the trace,and used Space Syntax to calculate the global integration (Rn) and spatial depth.Then,correlation analysis and multiple linear regression model by SPSS were used to research on the pedestrian distribution under the effect of multi-factors of space in rail transit complex.The main findings are summarized as following : (1) Relationship with the building entrance is proved to have greatest influence on pedestrian distribution.The entrances are arranged on the multiple floors in Langham Place,that helps to guide pedestrian distribution effectively.(2) The new combined variable I * F,spatial connection coefficient of function,also plays a major role in prediction of pedestrian distribution,which can remedy the limitation of space syntax without considering the effect of function.In intensive pattern of rail transit complex,in order to realize an acceptable pedestrian distribution on different floors,the functions should be arranged on the basis of the spatial integration.(3) The distance to escalators also has great influence on pedestrian distribution.On the same floor,the passage way connecting to escalator usually has high pedestrian flow than the one dose not.After revised by the variable distance to escalators,the predictive ability can be improved (adjusted R-square rising from 0.586 to 0.610).(4) The arrangement of atrium,multi-level escalators and spiraling retail-lined path in Langham Place,successfully guide distribution of the customers in vertical.(5) No correlation exists between relationship with atrium and pedestrian distribution,and the variable relationship with atriums is also excluded from the predicting model.To a certain extent,the reasons lie in the spatial organization patterns of Langham Place.