Flavonoids are a large family of plant and fungal natural products, among which many have been found to possess outstanding biological activities. Utilization of engineered microbes as surrogate hosts for heterologous biosynthesis of flavonoids has been investigated extensively. However, current microbial biosynthesis strategies mostly rely on using one microbial strain to accommodate the long and complicated flavonoid pathways, which presents a major challenge for production optimization. Here, we adapt the emerging modular co-culture engineering approach to rationally design, establish and optimize an Escherichia coli co-culture for de novo biosynthesis of flavonoid sakuranetin from simple carbon substrate glucose. Specifically, two E. coli strains were employed to accommodate the sakuranetin biosynthesis pathway. The upstream strain was engineered for pathway intermediate p-coumaric acid production, whereas the downstream strain converted p-coumaric acid to sakuranetin. Through step-wise optimization of the co-culture system, we were able to produce 29.7 mg/L sakuranetin from 5 g/L glucose within 48 h, which is significantly higher than the production by the conventional monoculture-based approach. The co-culture biosynthesis was successfully scaled up in a fed-batch bioreactor, resulting in the production of 79.0 mg/L sakuranetin. To our knowledge, this is the highest bioproduction concentration reported so far for de novo sakuranetin biosynthesis using the heterologous host E. coli. The findings of this work expand the applicability of modular co-culture engineering for addressing the challenges associated with heterologous biosynthesis of complex natural products. KEY POINTS: • De novo biosynthesis of sakuranetin was achieved using E. coli-E. coli co-cultures. • Sakuranetin production by co-cultures was significantly higher than the mono-culture controls. • The co-culture system was optimized by multiple metabolic engineering strategies. • The co-culture biosynthesis was scaled up in fed-batch bioreactor.