Since 2013, North America has experienced a sharp increase in unintentional fatal overdoses: fentanyl, and its analogues, are believed to be primarily responsible. Currently, the most practical means for people who use drugs (PWUD) to avoid or mitigate risk of fentanyl-related overdose is to use drugs in the presence of someone who is in possession of, and experienced using, naloxone. Self-test strips which detect fentanyl, and some of its analogues, have been developed for off-label use allowing PWUD to test their drugs prior to consumption. We review the evidence on the off-label sensitivity and specificity of fentanyl test strips, and query whether the accuracy of fentanyl test strips might be mediated according to situated practices of use. We draw attention to the weak research evidence informing the use of fentanyl self-testing strips.