Deficiency of ectodysplasin A1 (EDA1) due to variants of the gene causes X-linked hypohidrotic ectodermal dysplasia (XLHED), a rare genetic condition characterized by abnormal development of ectodermal structures. XLHED is defined by the triad of hypotrichosis, hypo- or anhidrosis, and hypo- or anodontia. Anhidrosis may lead to life-threatening hyperthermia. A definite genetic diagnosis is, thus, important for the patients' management and amenability to a novel prenatal treatment option. Here, we describe five familial EDA variants segregating with the disease in three families, for which different prediction tools yielded discordant results with respect to their significance. Functional properties and levels of circulating serum EDA were compared with phenotypic data on skin, hair, eyes, teeth, and sweat glands. EDA1-Gly176Val, although associated with relevant hypohidrosis, still bound to the EDA receptor (EDAR). Subjects with EDA1-Pro389LeufsX27, -Ter392GlnfsX30, -Ser125Cys, and an EDA1 splice variant (c.924+7A > G) showed complete absence of pilocarpine-induced sweating. EDA1-Pro389LeufsX27 was incapable of binding to EDAR and undetectable in serum. EDA1-Ter392GlnfsX30, produced in much lower amounts than wild-type EDA1, could still bind to EDAR, and so did EDA1-Ser125Cys that was, however, undetectable in serum. The splice variant c.924+7A > G resulted experimentally in a mix of wild-type EDA1 and EDA molecules truncated in the middle of the receptor-binding domain, with reduced EDA serum concentration. Thus, assays reflected the clinical phenotype in two of these difficult cases, but underestimated it in three others. Absence of circulating EDA seems to predict the full-blown phenotype of XLHED, while residual EDA levels may also be found in anhidrotic patients. This indicates that unborn subjects carrying variants of uncertain significance could benefit from an upcoming prenatal medical treatment even if circulating EDA levels or tests suggest residual EDA1 activity.