<p>Smoke inhalation is a leading cause of fire-related deaths, primarily due to toxic gases like carbon monoxide (CO) and hydrogen cyanide (HCN). Both gases often coexist in fire victims, but while CO poisoning is commonly recognized in emergency settings, HCN intoxication is often underdiagnosed, partly due to the lack of rapid diagnostic tools and the overlapping clinical presentation with CO poisoning. Diagnosis must therefore rely on presumptive clinical suspicion based on clinical and laboratory findings. Early recognition is crucial for initiating appropriate treatment. The objective of the study was to describe the prevalence and clinical characteristics of presumptive HCN poisoning among fire victims, using clinical and analytical criteria. This study also aims to evaluate the overlap with CO poisoning and discuss surrogate markers that may aid in clinical suspicion. A retrospective observational study over 10&#xa0;years was conducted at a tertiary-care hospital emergency department. Fire victims were identified through electronic records, and those with arterial blood gas analysis were included. Presumptive HCN diagnosis was established according to predefined criteria, including the presence of neurological or cardiovascular symptoms, carboxyhemoglobin &gt; 15%, or lactate &gt; 7.5&#xa0;mmol/L. These criteria, while based on literature, are also commonly associated with severe CO poisoning. Therefore, the diagnosis of HCN poisoning remains non-confirmatory. Out of 172 patients, 26.2% had CO poisoning and 8.1% met criteria for presumptive HCN poisoning, all of whom also had CO intoxication. The presence of nasal soot was associated with a high negative predictive value, but failed to identify positive cases, suggesting its utility primarily as a tool to rule out HCN intoxication. Neurological symptoms were common, but frequently unrecognized. Only 42.9% of HCN intoxications were correctly identified in the emergency department, with a significant number of cases missed, particularly those with neurological signs. No patients had lactate &gt; 7.5&#xa0;mmol/L, although HCN-intoxicated patients showed higher mean levels than other groups. Among the correctly diagnosed, all received hydroxocobalamin, while undiagnosed patients did not. One patient, who was not given hydroxocobalamin, died with concurrent presumptive HCN poisoning. It is unknown whether this contributed to his death. HCN poisoning remains underrecognized in fire victims. Because confirmatory testing is lacking, diagnosis must be considered presumptive based on predefined criteria. Clinical signs traditionally used for CO severity may also suggest possible HCN exposure, but their specificity is limited and must be interpreted with caution. The absence of airway soot may help rule out HCN poisoning. Our findings underscore the importance of clinical vigilance in the absence of confirmatory testing.</p>