To compare subjective and objective measures of adherence to prophylaxis in haemophilia. In this cross-sectional study, we compared participants' self-perceived adherence and their estimate of the number of clotting factor concentrates (CFCs) that had been missed over the last period of CFC dispensation with an objective measure of adherence based on counts of CFC vials returned by participants. We included 29 out of 31 eligible patients in the study. There was no significant correlation between self-perceived degree of adherence and the objective classification of adherence (Rho: 0.10, 95% CI: -028 to 0.46, P: 0.61) and between the classification of adherence based on the proportion of missed CFC doses assessed by participants' self-report and objectively (Rho: 0.32, 95% CI: -0.01 to 0.59, P: 0.11). Conversely, we found evidence of moderate correlation between the proportion of missed CFC doses as assessed by participants' self-report and objectively (Rho: 0.56, 95% CI: 0.24 to 0.77, P: 0.003). Participants' self-perceived adherence was 3 times more likely to be rated as very good or good than it was for the objective assessment to be classified as adherent or suboptimally adherent. Our results showed significant discrepancies between subjective and objective measures of adherence, which likely reflect the influence of social desirability bias in self-reported measures and different concepts of adherence between patients/caregivers and haemophilia experts. Additionally, our results allow us to hypothesize that studies on adherence to prophylaxis in haemophilia relying exclusively on information from self-reports and questionnaires may substantially overestimate adherence levels. © 2019 John Wiley & Sons Ltd.