In Tanzania, diets are dominated by starchy staple crops such as maize, levels of malnutrition are high and largely attributed to lack of dietary diversity. We employed fuzzy cognitive mapping to understand the current soybean, maize and chicken value chains, to highlight stakeholder relationships and to identify entry points for value chain integration to support nutritious diets in Tanzania. The fuzzy cognitive maps were constructed based on information gathered during household interviews with 569 farming households, followed by a participatory workshop with 54 stakeholders involved in the three value chains. We found that the soybean, maize and chicken value chains were interconnected, particularly at the level of the smallholder farming systems and at processing facilities. Smallholder farming households were part of one or more value chains. Chicken feed is an important entry point for integrating the three value chains, as maize and soybean meal are the main sources of energy and protein for chicken. Unlike maize, the utilization of soybean in chicken feed is limited, mainly due to inadequate quality of processing of soybean grain into meal. As a result, the soybean grain produced by smallholders is mainly exported to neighbouring countries for further processing, and soybean meal is imported at relatively high prices. Enhancing local sourcing and adequate processing of soybean, coupled with strengthening the integration of smallholder farmers with other soybean, maize and chicken value chain actors offers an important opportunity to improve access to nutritious diets for local people. Our method revealed the importance of interlinkages that integrate the value chains into a network within domestic markets.