The benefits of exercise and behavioural recommendations in gestational diabetes mellitus (GDM) are controversial. In a randomized trial with a 2×2 factorial design, we examined the effect of exercise and behavioural recommendations on metabolic variables, and maternal/neonatal outcomes in 200 GDM patients. All women were given the same diet: group D received dietary recommendations only; group E was advised to briskly walk 20‐min/day; group B received behavioural dietary recommendations; group BE was prescribed the same as B + E. Dietary habits improved in all groups. In a multivariable regression model, fasting glucose did not change. Exercise, but not behavioural recommendations, was associated with the reduction of postprandial glucose (p < 0001), glycated haemoglobin (HbA1c; p < 0.001), triglycerides (p = 0.02) and C‐reactive protein (CRP; p < 0.001) and reduced any maternal/neonatal complications (OR = 0.50; 95%CI=0.28–0.89;p = 0.02). In GDM patients a simple exercise programme reduced maternal postprandial glucose, HbA1c, CRP, triglycerides and any maternal/neonatal complications, but not fasting glucose values.