BACKGROUNDThe black soldier fly (BSF) offers a potential solution to address the ongoing depletion of food/feed source, yet selecting effective rearing substrates remains a major hurdle in BSF farming. In an urban city, like Singapore, much of the current practices are based on rearing BSF on homogeneous waste streams (e.g., spent brewery grains, okara, etc.) because heterogeneous food wastes (e.g., mixed kitchen/canteen waste, surplus cooked food, etc.) present several operational challenges with respect to standardizing development, nutritional content, and harvesting.RESULTSIn this study, we compared two genetic strains (wild type and laboratory‐adapted line) of BSF larvae in a bioconversion experiment with diverse types of food waste (homogeneous/heterogeneous; plant/meat) and quantified the phenotypic plasticity. Our results demonstrate the different plasticity in bioconversion performance, larval growth and larval nutrition between the two BSF lines. This difference may be attributed to the selective breeding the laboratory‐adapted line has experienced. Notably, larval lipid content displayed little to no genetic variation for plasticity compared with larval protein and carbohydrate content. Despite variation in larval development, heterogeneous food wastes can produce better performance in bioconversion, larval growth, and larval nutrient content as compared to homogeneous food waste. All‐meat diets result in high larvae mortality, but larval survival could be rescued by mixing meat with plant‐based food wastes.CONCLUSIONOverall, we suggest using mixed meals for BSF larvae feeding. While targeted breeding may be a promising strategy for the BSF industry, it is important to note that selection effects on plasticity in larval nutrition should be carefully considered.This article is protected by copyright. All rights reserved.