The datasets in this archive contain all processed data of which the results are described in ‘No evidence for testosterone’s causal effect on decision-making under risk and ambiguity: A pre-registered triple-blind single-dose administration study in females’ (Woyke, Ikink, Heuvelmans, Roelofs, & Figner, Hormones and Behavior, 2020, forthcoming). The paper investigates the causal effect of testosterone on ambiguous and risky choice in the gain and loss domain, using a triple-blind, placebo-controlled, randomized, between-subjects design with 80 healthy female participants (18-27 years; M=21.4, SD=2.1). All analyses were pre-registered, and both researchers and reviewers remained blind to the treatment conditions (testosterone/placebo) until the manuscript was revised for publication. Participants received 0.5mg of testosterone or a matched placebo using a well-established, oral administration procedure, and completed 3.5h to 4.5h later a shortened version of the Risk and Ambiguity Task (RAT) described by Tymula et al. (2012, 2013). The documentation and codebook file describes the content of the datasets.