The purpose of this paper is to analyze the relationship between fluctuation of time-headaway and traffic flow conditions. Traffic data used in this investigation were observed at a two-lane street in the morning peak hour. Traffic flow are classified based on the critical speed of 30km/h into four types of traffic flow condition, i. e., non-congested flow, transitional flow from non-congested to congested flow, congested flow and transitional flow from congested to non-congested flow, and the relationship between changes of speed and standard deviation of time-headway are analyzed based on the observed time-series data. As the results, it becomes known that the extremely large standard deviations of headaway distribution aggregated by 3-minutes time duration appear at the period when traffic flow changes from one type of traffic condition to another. According to this result, the criteria to judge traffic condition are proposed based on the standard deviation of time-headway.