BACKGROUND: Survival from cardiac arrest is sensitive to the quality of delivered CPR. In 2010, updated international resuscitation guidelines emphasized deeper chest compressions and faster rates, yet it is unknown whether training laypersons using updated guidelines resulted in changed CPR performance. We hypothesized that laypersons taught CPR using the 2010 guidelines performed deeper and faster compressions than those taught using the 2005 materials. METHODS: This work represents a secondary analysis of a study conducted at eight hospitals where family members of hospitalized cardiac patients were trained in CPR. An initial cohort was trained using the 2005 guidelines, and a subsequent cohort was trained using the 2010 guideline materials. Post training, CPR skills were quantified using a recording manikin. RESULTS: Between May 2009 to August 2013, 338 subjects completed the assessment. Among the subjects, 176 received 2005 training and 162 underwent 2010 training. The mean compression rate in the 2005 cohort was 87 (95％CI 83–90) per minute, and in the 2010 cohort was 86 (95％CI 83–90) per minute (P=ns), while the mean compression depth was 34 (95％CI 32–35) mm in the 2005 cohort and 46 (95％CI 44–47) mm in the 2010 cohort (P<0.01). CONCLUSIONS: Training with the 2010 CPR guidelines resulted in a statistically significant increase in trainees' compression depth but there was no change in compression rate. Nevertheless, the majority of CPR performed by trainees in both cohorts was below the guideline recommendation, highlighting an important gap between training goals and trainee performance.