Introduction: Tuscany is one of the few Italian Regions implementing extensively the organisational model of Primary Care Centres (PCCs), based on the co-location of General Practitioners within the same structure with nurses, specialists, social workers and administrative staff and other primary and secondary care services. Although co-location of professionals is considered a key factor for integrated care, it is not sufficient per se to ensure inter-professional collaboration. The aim of the study was to analyse how the professional and clinical integration, in particular the collaboration between professionals of different disciplines, is characterised within PCCs. Methods: Three cases studies of PCCs in Tuscany Region were carried out. Data collection was made through a web survey involving 107 professionals. Social Network Analysis (SNA) was applied. The analysis included visual representation of the networks diagrams, analysis of the network structure, calculation of key network measures and possible associations among them. Results: The response rate differs among the three PCCs and it gives indirectly a proxy measure of the different level of professional involvement in PCC activities. The sharing of electronic data involves the overall networks, but it is mainly functioning among professionals with the same profile and among professionals working within the same service. There are different extents of clinical integration among all professionals. In all PCCs the core nodes of the clinical integration are GPs and nurses, while specialists and social workers are differently involved in the patient care management. Discussion: This study highlights some potentialities and limits of the actual inter-professional collaboration in PCCs. Functional integration, based on sharing electronic patient data, is quite spread although it is more clustered by different disciplines and services. Clinical integration is quite strong between GPs and nurses. The involvement of social workers in patient care pathway should be improved. Sometimes specialists remain as a separate subgroup within the general network of professionals, both as informal relationships and professional relationships. Conclusion: This study gives an innovative perspective on the inter-professional integration within PCCs and provide suggestions supporting decision makers and health services managers to improve professional integration. Lesson learned: The success in implementing the organisational model of PCCs for a better quality of care will depend significantly on their real capacity to promote inter-professional collaboration among the professionals working in these structures. Formalisation of multidisciplinary teams, with defined role and responsibilities, inter-professional and team building training, strengthening of a common information system among professionals may contribute to improve integration in PCCs. Limitations: The non-respondents might have conditioned an under-representation of existing relationships among professionals working in the structure analysed. Since data were collected through questionnaire, some professionals could not have answered exhaustively because of confidentiality reasons, although the closed-ended question format reduce greatly the risk of recall bias. Suggestions for future research: SNA could be applied also in investigating the characteristics of singles nodes within the network (in terms of their position and relationships with the other nodes) and to correlate these measures with the individual perceptions of the PCCs results.