BACKGROUND: We aimed at identifying the factors influencing the natural history of non-IgE-mediated gastrointestinal food allergies (non-IgE-GIFA), a group of common paediatric conditions including food protein-induced: enteropathy (FPE), allergic proctocolitis (FPIAP), enterocolitis syndrome (FPIES), and motility disorders (FPIMD). METHODS: Prospective multicentre cohort study involving paediatric patients (both sexes, aged ≤14 y) with non-IgE-GIFA diagnosed and followed for 24 months at a Tertiary Centre for Paediatric Allergy, Gastroenterology and Nutrition. Anamnestic and clinical data were collected from all enrolled patients. RESULTS: 123 non-IgE-GIFA patients were enrolled (56% male, median age (IQR) 150 (60-300) days): FPE (39%), FPIES (17%), FPIAP (16%) and FPIMD (28%). 42% of patients had multiple food allergies (FAs) at baseline, and 64% had a positive family history of allergy. Male sex (OR = 2.24, 95% CI 1.07 to 4.71) and every 1 month of diagnostic delay (OR=1.09, 95% CI 1.01 to 1.18) were positively associated with the occurrence of multiple FAs. At 24-month follow-up, 54% of patients acquired immune tolerance. This rate was higher in FPIAP (75%), when compared with FPIMD (62%), FPE (54%) and FPIES (24%). The odds of 24-month immune tolerance acquisition rate was lower in children with family history of allergy (OR=0.41, 95% CI 0.19 to 0.89) and in those with multiple FAs at baseline (OR=0.24, 95% CI 0.11 to 0.51). At 24-month follow-up, the rate of patients with allergic march was 0.46 (95% CI 0.38 to 0.55, n=57/123), without differences comparing the four phenotypes. The presence of multiple FAs at baseline was associated with an increased risk of developing allergic march (OR=2.22, 95% CI 1.07 to 4.61) at 24-month follow-up. CONCLUSIONS: The results of the study suggest the potential role of modifiable and non-modifiable risk factors influencing the natural history of paediatric patients affected by non-IgE-GIFA.