As we are all aware, acute invasive fungal rhinosinusitis (AIFR) is a life-threatening disease which typically affects immunocompromised patients. It is diagnosed via typical signs, symptoms and presentation but gold standard is by biopsy of tissues demonstrating infiltration fungal hyphae. Treatment is surgical debridement of all necrotic tissue and systemic antifungals. Mortality rate remains high at 30–80%, and poor prognosis is associated with delayed surgical intervention, bilateral disease and extra sinus extension. Early diagnosis is key but challenging and the authors seek to develop a statistical predictive model to effectively differentiate AIFR from other types of sinonasal diseases. The authors looked at 135 high-risk AIFR patients, 67 of whom had biopsy-proven AIFR over a 15-year period. Initially, the authors proposed four models which looked at signs and symptoms, symptoms and endoscopy, CT imaging and all three criteria as one. The model with the most predictive value was the latter, with a sensitivity of 74.63% and specificity of 89.71%. The authors define two major criteria – visual loss; mucosal discolouration, (the presence of either of these scores two points each) – and four minor criteria of fever, nasal crusting, mucosal loss of contrast or retro-antral fat stranding on CT (each of which causes one point). They considered a result positive when either two major criteria, or one major and one minor, or three minor criteria were present. This is a useful approach and could aid earlier detection of this devastating disease and, therefore, prompt a more efficient treatment.