Giardia duodenalis is a widespread intestinal protozoan that affects mammals, including humans. Symptoms can range from being subclinical to causing severe abdominal pain and diarrhoea. Giardiasis often requires repeated treatment with synthetic drugs like metronidazole. In recent years, treatment failures in clinical cases involving nitroimidazoles have been increasingly reported. Consequently, identifying therapeutic alternatives is necessary. Medicinal plants have traditionally been used as antiparasitic compounds, but systematic evaluation under controlled experimental conditions is often lacking. Here, we evaluated the in vitro efficacy of Tabebuia avellanedae dry and hydroalcoholic extracts, as well as one of its active compounds, β-lapachone, as potential treatment against G. duodenalis infection. We observed effective antigiardial activity for all tested compounds, with β-lapachone exhibiting lower IC50 values than metronidazole. Cytotoxic effects often limit therapeutic concentration windows of opportunity, and choosing an informative model to assess them is not straightforward. In the present case, only T. avellanedae hydroalcoholic extract showed no cytotoxicity on tumoral human intestinal Caco-2 cell line, and only a trend of inhibition when tested on canine epithelial kidney MDCK cells. To introduce a more physiological test system, we used in vitro G. duodenalis infection experiments in a trans-well set-up using organoid derived monolayers (ODM) to assess at the same time drug efficacy against the parasite and safety on primary human intestinal epithelia, a likely surrogate for in vivo conditions. Our studies using this model point towards the potential therapeutic opportunity for non-systemic applications of T. avellanedae extracts and a relevant ingredient of these, β-lapachone. The data suggest that ODM co-cultures with G. duodenalis are suitable for testing antigiardial compounds, providing a more informative in vitro model before progressing to in vivo tests.