In COVID-19 patients, procalcitonin (PCT) and C-reactive protein (CRP) performance in identifying bacterial infections remains unclear. Our study aimed to evaluate the association of PCT and CRP with secondary infections acquired during ICU stay in critically ill COVID-19 patients. This observational study included adult patients admitted to three COVID-19 intensive care units (ICUs) from February 2020 to May 2022 with respiratory failure caused by SARS-CoV-2 infection and ICU stay ≥ 11 days. The values of PCT and CRP collected on the day of infection diagnosis were compared to those collected on day 11 after ICU admission, the median time for infection occurrence, in patients without secondary infection. The receiver operating characteristic curve (ROC) and multivariate logistic model were used to assess PCT and CRP association with secondary infections. Two hundred and seventy-nine patients were included, of whom 169 (60.6%) developed secondary infection after ICU admission. The PCT and CRP values observed on the day of the infection diagnosis were larger (p < 0.001) than those observed on day 11 after ICU admission in patients without secondary infections. The ROC analysis calculated an AUC of 0.744 (95%CI 0.685–0.803) and 0.754 (95%CI 0.695–0.812) for PCT and CRP, respectively. Multivariate logistic models showed that PCT ≥ 0.16 ng/mL and CRP ≥ 1.35 mg/dL were associated (p < 0.001) with infections acquired during ICU stay. Our results indicated that in COVID-19 patients, PCT and CRP values were associated with infections acquired during the ICU stay and can be used to support, together with clinical signs, rather than predict or rule out, the diagnosis of these infections.