The 40-Something RCT aimed to determine if a 12-month health professional-led intervention could modify diet and physical activity behaviour for obesity prevention, in 44-50 year old, non-obese (BMI=18.5-29.9kg/m2) premenopausal women. Women were monitored for an additional 12 months to determine if effects could be maintained. This paper aimed to explore dietary and physical activity behavioural mediators hypothesized to be causally associated with weight change. Fifty-four women were randomised to a Motivational Interviewing Intervention (MI) (n=28; five health professional consultations) or a Self-Directed Intervention (n=26; written advice). Compliance to 10 study recommendations was measured at three months by a four-day weighed food and physical activity record including pedometer-measured step counts, self-reported exercise minutes and sitting time. The 10 compliance scores were independently assessed in mediation models for 12- and 24-month weight change. The MI effect on step count was an increase of 0.99 points on the 10-point compliance scale (p≤0.01). This MI effect on step count significantly mediated the 12 and 24 month effect on weight (12 months AB=-0.74, 95%CI=-1.95, -0.14; 24 months AB=-1.06, 95%CI=-2.56, -0.36), accounting for 37.23% and 53.79% of the effect, respectively. The MI effect on vegetable serves was an increase of 1.50 points on the compliance scale (p=0.02). The MI effect on vegetable compliance significantly mediated the effect on weight at 24 months (AB=-0.54, 95%CI=-1.50, -0.04), accounting for 24.92% of the effect. The remaining eight dietary and physical activity compliance scores did not significantly mediate weight loss. Encouraging women to take 10,000 steps and eat five vegetable serves per day may be a promising strategy to achieve long-term weight control at mid-life.