Most pre-eclampsia (PE) studies have used cross-sectional data to derive conclusions regarding the pathophysiology of the condition. This has led to the concept that there exists early (<34 weeks) and late-onset (>34 weeks) disease according to gestational age at diagnosis. Survival time models have predicted that if the pregnancy was to continue indefinitely, all women would develop PE. In this study we have performed a longitudinal analysis of the urinary biomarker, inositol phosphoglycan (IPG), in a cohort of women giving birth in Mauritius (n-920). We have analysed the PE data in the traditional cross-sectional manner for n = 77 women who developed PE and also then looked at the longitudinal data for 71/77 of the same women. The data allows us to use longitudinal values to calculate a date of onset (first presence of biomarker in urine) and compare that to date of clinical diagnosis (cross sectional). We find two populations for both analysis consistent with an early and late stage subgroup. The calculated date of onset had subgroups (early and late) at 28.4 ± 0.41 weeks and 35.37 ± 0.26 weeks and for clinical date of diagnosis, 32.3 ± 0.59 weeks and 37.04 ± 0.62 weeks, respectively. The presence of the same biomarker in both subgroups and its ability to predict clinical onset 2–4 weeks prior to clinical diagnosis suggest that both groups may have similar aetiology.Urinary inositol phosphoglycan PPreeclampsia