OBJECTIVE: to analyse women's experiences of early labour care in caseload midwifery in Australia. DESIGN: this study sits within a multi-site randomised controlled trial of caseload midwifery versus standard care. Participant surveys were conducted at 6-weeks and 6-months after birth. Free-text responses about experiences of care were subject to critical thematic analysis in NVivo 11 software. SETTING: two urban Australian hospitals in different states. PARTICIPANTS: women 18 years and over, with a singleton pregnancy, less than 24 weeks' pregnant, not planning a caesarean section or already booked with a care provider; were eligible to participate in the trial. INTERVENTIONS: participants were randomised to caseload midwifery or standard care for antenatal, labour and birth and postpartum care. MEASUREMENTS AND FINDINGS: The 6-week survey response rate was 58% (n = 1,019). The survey included five open questions about women's experiences of pregnancy, labour and birth, and postnatal care. Nine-hundred and one respondents (88%) provided free text comments which were coded to generate 10 categories. The category of early labour contained data from 84 individual participants (caseload care n = 44; standard care n = 40). Descriptive themes were: (1) needing permission; (2) doing the 'wrong' thing; and (3) being dismissed. Analytic themes were: (1) Seeking: women wanting to be "close to those who know what's going on"; and (2) Shielding: midwives defending resources and normal birth. KEY CONCLUSIONS: Regardless of model of care, early labour care was primarily described in negative terms. This could be attributed to reporting bias, because women who were neutral about early labour care may not comment. Nevertheless, the findings demonstrate a gap in knowledge about early labour care in caseload midwifery models. IMPLICATIONS FOR PRACTICE: Maternity services that offer caseload midwifery are ideally placed to evaluate how early labour home visiting impacts women's experiences of early labour.