We provide three data sets that were used to estimate bird habitat preference in an urban environment. The Phi-bird preferences and Logistic-bird preference data sets contain the necessary information to estimate bird habitat preference via the Phi association coefficient and logistic regression, respectively. Additionally, the estimates obtained with each approach are presented. In the Phi-bird preferences data set, the bird records collected from unstructured citizen science data and structured scientific information are shown. Finally, the Validation data set presents all the information used to validate the reliability of both, the Phi association coefficient and the Logistic regression methods