To analyse the association between adherence to anti-diabetic therapy (diet, physical activity and medications) and perinatal outcomes. Methods: A cohort design was used. Participants were 157 pregnant women with diabetes, and the setting was Mbuya Nehanda and Chitungwiza Maternity Hospitals, Harare, Zimbabwe. Results: Main outcome measures were maternal and perinatal outcomes. Mean adherence to anti-diabetic therapy was 66.7%. Perinatal outcomes observed were hypertensive disorders (34.5%), Caesarean delivery (45.9%), maternal diabetic ketoacidosis (5.1%), maternal hypoglycaemia (15.9%), and candidiasis (19.7%). Neonatal outcomes were perinatal mortality (15.9%), low Apgar score at 1 minute (26.8%), low Apgar score at 5 minutes (24.8%), macrosomia (33.8%), neonatal hypoglycaemia (15.3%), and neonatal hyperbilirubinemia (7.6%). There were significant associations between adherence and Caesarean delivery (RR 1.9, 95% CI 1.28 to 2.81, p = 0.0014), candidiasis (RR 3.95, 95% CI 1.65 to 9.47, p = 0.002), low Apgar score at 1 minute (RR 2.15, 95% CI 1.16 to 3.98, p = 0.015) and at 5 minutes (RR 1.95, 95% CI 1.03 to 3.69, p = 0.039), and perinatal mortality (RR 3.08, 95% CI 1.11 to 8.52, p = 0.018). Conclusions: Adherence to anti-diabetic therapy was sub-optimal and was associated with some adverse perinatal outcomes. Promotion of adherence, through routine individualised counselling, monitoring and assessment, is vital to minimise adverse outcomes.