Background: urinary incontinence carries major social burden and considerable costs for health care systems. Objective: the aim of this study was to investigate changes in continence status among a large cohort of older women, and to identify factors associated with incidence of incontinence in later life. Subjects: participants of the Australian Longitudinal Study of Women's Health (ALSWH), aged 70–75 years in 1996 and who have completed four health surveys over the past 10 years. Methods: continence status across four survey periods, spanning 9 years, were defined according to women's reports of ‘leaking urine’ at each survey. Generalised estimating equation (GEE) models were used in longitudinal analyses of the factors associated with changing continence status over time. Results: this study presents longitudinal data on the prevalence and incidence of incontinence from a large cohort of older women, over 9 years of follow-up. Over this time, 14.6% (95% CI 13.9–15.3) of the women in the study who had previously reported leaking urine ‘rarely’ or ‘never’ developed incontinence, and 27.2% (95% CI 26.2%, 28.3%) of women participating in Survey 4 (S4) in 2005 reported leaking urine ‘sometimes’ or ‘often’ at that survey, with women being twice as likely to report incontinence at S4 as they were 6 years earlier. Longitudinal models demonstrated the association between incontinence and dementia (P < 0.001; OR = 2.34; 95% CI 1.64, 3.34), dissatisfaction with physical ability (P < 0.001; OR = 1.70; 95% CI 1.52, 1.89), falls to the ground (P <0.001; OR = 1.23; 95% CI 1.13, 1.33), BMI (P < 0.001; OR = 2.18; 95% CI 1.70, 2.80 for obese), constipation (P < 0.001; OR 1.46; 95% CI 1.34–1.58), urinary tract infection (P < 0.001; OR 2.07; 95% CI 1.89–2.28), history of prolapse (P 0.001; OR = 1.53; 95% CI 1.35, 1.74) and prolapse repair (P = 0.002; OR = 1.23; 95% CI 1.08, 1.40). Stroke (P = 0.01), parity (P = 0.017) and hysterectomy (P = 0.026) and number of visits to the general practitioner (P = 0.040) were less strongly associated with incontinence in the final longitudinal model. Incontinence was not significantly associated with area of residence (P = 0.344), education (P = 0.768), smoking (P = 0.055), diabetes (P = 0.072), attending support groups (P = 0.464) or attending social groups (P = 0.022). Conclusion: strong associations between BMI, dysuria and constipation indicate key opportunities to prevent incontinence among older women.