Lactose, the main carbohydrate in cow's milk, may cause health problems for consumers with intolerance. Lactose determination in milk is hindered by the matrix complexity and lack of chromophore groups. Chromatography, volumetric, and spectrophotometric approaches involving chemical derivatization are time-consuming and require laborious sample preparation, which is incompatible with the high analytical demand. In this context, a novel approach is presented for lactose determination in milk exploiting smartphone-based digital-image photometry. It was based on a modification of the Benedict's method, involving formation of the violet Cu(I)/2,2'-biquinoline-4,4'-dicarboxylate (BCA) complex instead of the copper(I) oxide precipitate, aiming at improvement of sensitivity and precision. Sample pretreatment and analyte derivatization were performed in Eppendorf tubes with minimal reagent amounts and a smartphone camera was used for image acquisition under controlled conditions. Measurements were based on the RGB color system, taking channel G as the analytical response because of the complementarity with the color of the complex. Under the optimized conditions, the proposed procedure yielded a linear response up to 20 mg L-1 (r = 0.999), with a limit of detection of 1.5 mg L-1, which is compatible with determination of lactose in milk and dairy products categorized with low content of the sugar. The procedure takes less than 10 min, with a coefficient of variation of 3.0% (n = 12) and consumes as low as 160 μg Cu and 430 μg BCA per determination, thus being a more practical, fast, cost-effective, and environmental friendly analytical method.