Based on the premise of flood management, administration of rivers in Taiwan consists of building many water structures to protect the lives and properties of people. But the ecological equilibrium of the area may possibly be damaged. Thus, how to maintain the ecological instream flow and how to lessen the negative impact of human development on river ecology are topics that have to be further investigated. The relationships between hydraulic conditions and ecological habitats are discussed and compared with quantified assessment technology in this study. Acrossocheilus paradoxus was selected as the target species, and a hydraulic model HESRAS 3.0 along with the habitat model RHABSIM 2.2 was employed to estimate the ecological instream flow of the habitat. The Wu-chi river in Taiwan was chosen for the case study. This study finds some results. First, when flow increases, the habitat spaces increase while the combined suitability factor (CSF) also increase in a relatively lower rate. And by this reason, the majority of habitat spaces are of the "run" type while a few "pool" types also existed. When flow rate is more than 200cms, the "pool" type diminished. Second, In the river ecology, pool type is an important refuge and perch for fishes. For flows which are too fast in the run areas (such as section 66), Acrossocheilus paradoxus may possibly be unable to cross, resulting in habitat cut offs, and which may hamper survival of this fish species given the habitat conditions. Third, Local and international results indicate that ecological flows are described as single values but for the animals, there should be a range which changes can occur. Thus, this study considers the effect of variety of habitats and habitat to set the ecological base flow limits of the section under study and to maintain the ecological habitat diversity required. In addition, we have set the ecological base flow limits for the research section between 10-40 cms as most suitable for Acrossocheilus paradoxus survival. Finally, table 1 indicates that in 1997, a total of 288 days failed to meet the ecological base flow standard while the longest number of days which did not meet the ecological base flow standard is 125 days during 1994. These two years should be the worst years for the ecology; if the year where the most number of days meeting the ecological base flow limits are seen as the best year for the ecology, then 1991 is the best (250 days) and then followed by 2000 (245 days). Moreover, Table 1 shows that 1991 has the lowest longest number of days which does not meet the ecological base flow limits (26 days). Thus, 1991 is the best year for the ecology. (Table Presented).