Aim: Variations in catch per unit effort (CPUE) of hilsa (Tenualosa ilisha ) at different habitats (sanctuaries) of Bangladesh waters were observed to understand whether CPUE variations were due to differences in habitat complexity, environmental factors or both. Methodology: CPUE of hilsa data were collected from five different habitats in Bangladesh waters between July 2015 and June 2016. Various physico-chemical parameters were measured monthly and then examined for any relationship to CPUE. Results: The overall mean CPUE of hilsa from the investigated areas was estimated to be 7.93 ± 6.03 kg gillnet day. The monthly average values of CPUE was the highest in January 2016 amounting 11.63 ± 1.21 kg gillnet day and the lowest in March 2016 amounting 3.63 ± 0.83 kg gillnet day with significant differences among the months (p < 0.05). Gillnet mesh sizes influenced the capture of hilsa with the highest (9.18 kg gillnet day) and lowest (6.05 kg gillnet day) catch obtained from the smallest (3.0 cm) and largest (11.0 cm) mesh size, respectively. There was a lower degree of positive and negative correlation (p > 0.05) with environmental variables and CPUE. Multiple regression analysis tests also indicated a weak overall correlation (62%) between environmental variables and CPUE of hilsa. Interpretation: Results likely indicate that habitat structure was more important in determining CPUE of hilsa (highest in Padma sanctuary) compared to environmental factors at the tested sites. This study emphasizes the need to conserve habitats that are often used as important fishing grounds for hilsa to ensure their adequate recruitment and ultimately sustainable management of this fishery