This paper presents the psychometric development of a new observer-report research questionnaire for assessing aspects of an individual’s mindfulness that are noticeable to others. Items from five established self-report mindfulness scales were re-worded for observer-report, and 30 were endorsed for potential inclusion by an expert panel (n = 5). Factor analytic and item response theory models were used to test item and scale psychometrics with data (N = 494) in three independent samples. A nine-item, three-factor scale with good fit indices was determined (RMSEA = 0.04, CFI = 0.99, TLI = 0.99). The instrument provides an overall score for observed mindful behaviours and subscale scores for observed attentiveness, awareness and acceptance. Within-subject test–retest reliability was strong (ICC 0.91) and agreement between observed mindful behaviours and self-reported mindfulness was adequate (ICC 0.45). Validity tests showed concordance between the new scale and self-reported mindfulness within the selected nomological network constructs (emotional intelligence, empathy and avoidant behaviours). Responsiveness was indicated but unconfirmed in data from a randomized controlled trial of low-dose mindfulness training. The Observed Mindfulness Measure (OMM) is a quantitative instrument that can provide an additional data source to strengthen self-reported findings in mindfulness research. With some further refinement this new instrument can advance research into whether and how mindfulness training might make a noticeable difference in social and organizational domains.