The nutritive value of pollen blends purchased by bumble bee producers from beekeepers is a key factor of successful mass rearing. We have already devised a method for quality ranking of pollen diets. It was a 1-mo bioassay using queenless micro-colonies of three callow workers (Bombus terrestris L.; Apidae: Bombinae). In the current study, we tested three pollen diets through the micro-colony method and compared the results to the development of queenright colonies supplied with the same diets. The three diets were “A,” a spring blend with a protein rate of 15.7% and a dominance of Corylus avellana L.(46%) and Buxus sempervivens L. (35%); “B,” a summer assemblage with a lower crude protein content (14.4%), with a dominance of Helianthus annuus L.(37%), Zea mays L. (35%), and Poaceae (21%); and “C,” a 50/50 mixture by weight of A and B, with a crude protein content of 15.8%. In micro-colonies, the more sensitive parameter for diet ranking was the mean weight of a larva, whereas in queenright colonies, the discrimination between treatments was enabled by considering the body size of new queens (weight and length of the radial cell) and the slope of the sigmoidal curve of the pollen consumption at the inflection point reflecting brood growth. Both testing methods assigned the lower rank to diet B compared with diet A. Despite the closeness of the three diets in terms of nitrogen content, we concluded that micro-colonies were a good estimate of colony development when nutritive value of pollen was tested.