As the need for customization of medical devices has become increasingly clear, designers have responded to the challenge. Despite extensive literature on the importance of customization, few studies have examined its impact on such wearable devices as glucose monitors. Little is known about how user experience and situational context may influence how diabetics choose these technologies and adapt them to manage their conditions. In this paper, we present a co-design study to learn about people’s use and the possibility of them building their own T1D glucose monitors, a device that is used on a daily basis to support everyday care. In the first phase, we interviewed and conducted brainstorming sessions to identify the strengths and challenges of current T1D glucose monitors and identified ideas for improving the monitor design. Our first phase revealed that users face three main problems with their monitors: they pose a physical hindrance, do not provide privacy, and are not aesthetically pleasing. Our brainstorming sessions highlighted that to address some of these concerns, customization was key. Informed by our data and literature, in the second phase, we built and evaluated our prototype system, DiaFit, consisting of interactive physical modules for building customizable T1D monitors. Our prototype consists of multiple modular accessories that people can assemble together to build a monitor which can be worn on the wrist, finger, ear or neck. The modules provide visual, audio and haptic feedback about the glucose levels. Our evaluation shows that participants could build their own glucose monitors using DiaFit and found these devices could better support their need for form, privacy, and sense of aesthetics. Users could create combinations of feedback and different accessories depending on the situational context. In particular, adults who self-identified as makers were particularly interested in using DiaFit and building different versions of the monitors more often. We found that customization and a do-it-yourself (DIY) approach could prove useful in the development of inclusive designs for T1D. Our main contribution is our design and evaluation of a prototype system for building personalized wearable T1D monitors using a variety of modules and accessories in a DIY approach. We conclude by discussing lessons learned and outline implications for future researchers interested in exploring bespoke DIY solutions for medical devices.