Introduction: One in 6 adults has a mental health problem, and 1 in 100 have a severe mental illness (SMI)1. Premature mortality rates are higher in those with mental ill health and is attributable to a range of factors including higher prevalence of preventable risk factors for chronic disease, and poorer access to physical healthcare3,4 5. Recently in England, there has been a policy focus to achieve ‘parity of esteem’ for physical and mental health, but it is unclear if this aspiration is reflected in practice. The aims of this study were to examine levels of hospital activity for physical health needs in those with mental ill health in England, and examine differences compared to those without mental ill health, over time. Methods : We used hospital data for England (hospital episode statistics) to identify individuals aged 18 to 75 years with an indication of mental ill health in each year between 2009/10 and 2013/14. Those with mental ill health (MH Cohort) were defined as using hospital services with a primary diagnosis (ICD-10 codes ‘F’ chapter) or under then main specialty (starting with ‘7’) of mental health. SMI was defined as a primary diagnosis of schizophrenia, bipolar disorder or psychoses (ICD-10 codes: ‘F20-29’ and ‘F30-31’). A comparator cohort (PH cohort) were those who had not used hospital services for mental ill health in each year nor previous two years. Crude hospital activity rates were calculated for the MH and PH cohorts, from 2009/10 to 2013/14. Hospital activity relating to mental or physical health was identified using the main speciality recorded during contact with the hospital services. We also examined rates of common inpatient procedures for physical health conditions, and characteristics of these admissions, e.g. planned versus emergency care, and length of stay. Results : We identified 545,759 patients in the MH cohort and 12.5M in the PH cohort in 2013/14. The MH cohort were younger and more deprived. Using hospital data, 21.3% of MH cohort had at least one other long-term condition recorded (in addition to mental ill health) compared to 23.7% of PH cohort. In 2013/14, compared to the PH cohort, the MH and SMI cohorts had higher A&E attendance rates (RR 3.2 and 3.3 respectively), emergency inpatient admissions (RR 4.9 and 6.7), and emergency admission rates for ambulatory care sensitive conditions (MH cohort only RR 3.6). From 2009/10 to 2013/14, the difference in emergency care use between those with mental ill health and those without, increased. Deprivation was strongly associated with emergency care in both the MH and PH cohorts, with the more deprived people using the most emergency hospital care. In 2013/14, the MH cohort also had lower planned inpatient admissions (RR 0.9), but higher rates or outpatient activity (RR 1.7). There was no change in the difference in planned care over the study period. In the MH cohort, 80.9% of inpatient admissions, and 48.6% of outpatient appointments were for physical health conditions (2013/14). For some common physical procedures a higher proportion were carried out as an emergency in the MH cohort compared to the PH cohort (eg upper digestive tract procedures (21.0% and 4.9% respectively)), and had a longer length of stay (e.g. hip replacement mean length of stay 11.4 and 4.7 days respectively). Discussion : We found higher rates of emergency care use and potentially preventable admissions in those with mental ill health. We found that those with mental ill health were more likely to have an emergency rather than planned admission for common inpatient procedures, be admitted overnight and stay longer in hospital. Given the complex mental and physical health needs in this vulnerable population, high emergency hospital care may be expected. But in this study we found high levels of potentially avoidable admissions and high rates of emergency admissions for physical health needs, which may reflect poor identification/management of physical health in those with mental ill health. Conclusion : This study highlights consistent disparities in the use of hospital services for physical health care in those with mental ill health in England over the past 5 years. A key limitation is that the analysis only reflects acute care activity, and more could be understood by linking to other care settings (eg primary care). However, if the gap in hospital care between those with mental ill health and those without continues further, then the national goal to break down barriers between mental and physical health and achieve parity of esteem by 2020 is unlikely to be realised.