Objective: as part of an evaluation of a team midwifery scheme we assessed the satisfaction of community and hospital midwives and their views about working practices and care provided. Design: survey of complete enumeration of community midwives (most working in teams) and hospital midwives providing antenatal, intrapartum and postnatal care to a population of women. Setting: community and district general hospital, in the UK. Measurements: socio-demographic data about midwives, ratings on Likert-type scales of job satisfaction, quality of care variables, relationships with other professionals and women; Glasgow Midwifery Process Questionnaire. Findings: 80 out of 92 midwives (87%) responded. Community midwives were younger, more recently qualified, employed on lower grades, less likely to be married and have children than hospital midwives. The Glasgow Midwifery Process Questionnaire revealed that midwives, particularly hospital midwives, had low morale. Community midwives were more likely to report that their job was satisfying, offered a variety of work, enabled them to use skills and knowledge fully, and offered opportunities for professional development. Hospital midwives were more likely to report following strict guidelines. Community midwives, however, disliked the long on call and unsociable hours, and reported disruption to family/social life. Forty-one per cent of hospital midwives (12) and 28% of community midwives (14) reported regularly working beyond their shift. Whilst midwives thought that team midwifery was, in theory, a good idea, in practice it was not working well because of the size of teams and caseload. About half the community midwives felt that teams had detrimentally affected the quality and continuity of care. Conclusions: whilst team midwifery aims to improve continuity of maternity care, in this instance, it does not appear to achieve this aim. Many midwives reported it had adversely affected care. Team midwifery is a source of disillusionment for midwives, since the continuity of carer ideal is unachievable in a system based on teams of seven or more. Attendance at the delivery may be a luxury provided at the expense of antenatal and postnatal continuity. Implications: midwives recommended remedial measures: reducing team sizes, reducing caseloads, ensuring teams were fully staffed, reducing ‘on call’ and labour ward hours. It remains to be seen whether these will have the desired effects on continuity of care.