The development of human behavior in fire into an area of scholarly study of vital importance has been extremely rapid. The advent of performance based fire safety regulations and codes together with the need for robust computer evacuation simulation models gives further impetus and sense of purpose for &re endeavors. Hard fire science alone cannot solve the "fire problem". With increasing international emphasis on community fire safety policy initiatives knowledge of occupant behavioral characteristics associated with fire is essential. To develop the human behavior knowledge base a coherent collaborative strategic research program which delivers value for money is essential. In this paper some thoughts on the way forward are presented for discussion.