Context: Achieving successful conservation outcomes in habitat creation and reintroductions requires an understanding of how species use their habitat and respond to these interventions. However, few initiatives directly compare microhabitat selection between remnant and managed habitats to measure effectiveness and evaluate outcomes. Probability of detection is also rarely included in studies on microhabitat use, which may lead to erroneous conclusions if detectability varies between variables. Methods: In this study, we used the endangered green and golden bell frog (Litoria aurea) to compare differences in microhabitat-use patterns in both a remnant and a constructed habitat. A detectability study was also conducted to determine detection probabilities among microhabitats. Key results: Aquatic vegetation was used more than expected in both the remnant and constructed habitats, and rock piles were utilised less than expected in the constructed habitat, despite their recommendation in most habitat templates. We found that detection probabilities altered the outcomes of abundance estimates for nearly all the measured microhabitat variables. Conclusions: Future management for this species should focus on providing high proportions of aquatic vegetation. Furthermore, although rock piles have been utilised greatly in past L. aurea habitat creation, placing large rocks on a managed site is expensive and time consuming. Future management initiatives may need to focus on providing smaller proportion of rocks, which would be a more appropriate use of resources. Implications: With conservation management projects increasing over the next few decades, understanding habitat use before implementing strategies should be a priority as it will provide important insights and inform decision-making for optimum habitat creation and restoration. Furthermore, accounting for detectability in microhabitat use studies is essential to avoid wrong conclusions that may negatively affect the success of ecological management strategies.