The aim of the study was to evaluate the food intake of pregnant women with gestational diabetes mellitus (GDM) according to two methods of dietary guidance. A randomised controlled clinical trial was conducted by appointment with a nutritionist and by using data from hospital records (2011–2014). The study population comprised adult women diagnosed with GDM treated in a public maternity hospital in Rio de Janeiro, Brazil. The control group (CG) received nutritional advice by the traditional method and the intervention group (IG) were instructed on carbohydrate counting. The analysis of food intake and the consumption of processed foods (PF) and ultra-processed foods (UPF) were evaluated in the second and third trimester. A total of 286 pregnant women were initially assessed (145 in the CG and 141 in the IG). It was observed that 89/120 (74·2 %) and 183/229 (79·9 %) consumed PF daily in the second and third trimesters, respectively, whereas 117/120 (97·5 %) and 225/231 (97·4 %) consumed UPF daily in the second and third trimesters, respectively. When analysing the intake of macronutrients (%) by quartiles, women who had fat intake in the third quartile had the highest average postprandial blood glucose compared with those who consumed fat in the second quartile (P=0·02). The consumption of PF and UPF was high and dietary intake was similar in both groups, regardless of dietary guidance method deployed, suggesting that both methods tested in the study can be used for monitoring the nutritional status of pregnant women with GDM.