Objective Compared with no music (NM), does listening to 'Achy breaky heart' (ABH) or 'Disco science' (DS) increase the proportion of prehospital professionals delivering chest compressions at 2010 guideline-compliant rates of 100-120 bpm and 50-60 mm depths? Methods A randomised crossover trial recruiting at an Australian ambulance conference. Volunteers performed three 1-min sequences of continuous chest compressions on a manikin accompanied by NM, repeated choruses of ABH and DS, prerandomised for order. Results 37 of 74 participants were men; median age 37 years; 61% were paramedics, 20% students and 19% other health professionals. 54% had taken cardiopulmonary resuscitation training within 1 year. Differences in compression rate (mode, IQR) were significant for NM (105, 99-116) versus ABH (120, 107-120) and DS (104, 103-107) versus ABH (p<0.001) but not NM versus DS (p=0.478). Differences in proportions of participants compressing at 100-120 bpm were significant for DS (61/74, 82%) versus NM (48/74, 65%, p=0.007) and DS versus ABH (47/74, 64%, p=0.007) but not NM versus ABH (p=1). Differences in compression depth were significant for NM (48 mm, 46-59 mm) versus DS (54 mm, 44-58 mm, p=0.042) but not NM versus ABH (54 mm, 43-59 mm, p=0.065) and DS versus ABH (p=0.879). Differences in proportions of subjects compressing at 50-60 mm were not significant (NM 31/74 (42%); ABH 32/74 (43%); DS 29/74 (39%); all p>0.5). Conclusions Listening to DS significantly increased the proportion of prehospital professionals compressing at 2010 guideline-compliant rates. Regardless of intervention more than half gave compressions that were too shallow. Alternative audible feedback mechanisms may be more effective.