This retrospective study evaluated energy and nutrient intake of dogs in a weight loss programme. Ninety‐four obese dogs were divided into three groups: G5–15: from 5% to 15% body weight (BW) loss (n = 55); G155–25: from 15.1% to 25% BW loss (n = 29); and G > 25: more than 25.1% BW loss (n = 10). Five brands of kibble diets designed for weight loss were analysed for crude protein, amino acids, fat, dietary fibre, and minerals. The food metabolizable energy (ME) was estimated (NRC, 2006). Data were compared inside each group using the paired t test and between groups with analysis of variance and Tukey tests (p < 0.05). The BW loss (weeks in regimen) was: G5–15, 9.8% ± 2.7% (13.5 ± 5.7 weeks); G15–25, 17.5% ± 2.7% (22.6 ± 11.9 weeks); G > 25, 30.0% ± 2.1% (50.4 ± 17.4% weeks; p < 0.01). The mean weekly BW loss rate was similar between groups (0.8% ± 0.3%; p > 0.05) but was higher during the first (0.96.6% ± 0.5%) than in the second (0.64% ± 0.4%) half of the regimen (p < 0.01). At beginning ME intake for BW loss did not differ (251.6 ± 32.2 kJ/kg0.75/day), but it was lower for G > 25 in the second half of the regimen (230.3 ± 44.3; p = 0.02). Considering the observed ME intake of each dog, depending on the commercial product, intake below recommended for maintenance was verified for crude protein in 1%–20% of the dogs, methionine in 4%–38% of dogs, methionine plus cystine in 4%–22% of dogs, tryptophan in 7%–93% of dogs, potassium in 2%–85% of dogs and magnesium in 1% up to 95% of dogs. Although the diets presented elevated nutrient concentrations per MJ, due to the reduced energy allowance, the estimated intake of several nutrients was lower than the recommendations, highlighting the importance of changing the formulation perspective, which must prioritize the actual nutrient intake per kg of BW during the energy deficit.