Aim To identify common drug-related problems (DRPs) during pharmacy intervention and consultation in an intensive care unit (ICU); to explore the gap between physicians and pharmacists on their understanding of each other’s capabilities and needs. Method We conducted a single-center prospective study in the ICU of a tertiary academic hospital for 21 months. A pharmaceutical care (PC) model was implemented by a pharmacy team, and data were collected during pharmacy intervention and consultation. Data analysis was performed on identified DRPs, causes and their relationships. DRPs’ frequency during intervention and consultation was compared. Problem-level descriptive analysis and network analysis were conducted using R 3.6.3. Result Implementation of PC model greatly improved the efficacy of pharmacists in both interventions proposed to solve DRPs (from 13.6 to 20.1 cases per month) and number of patients being closely monitored (from 7.7 to 16.9 per month). Pharmacists identified 427 DRPs during pharmacy intervention with primarily adverse drug events (ADEs, 34.7%) and effect of treatment not optimal (25.5%), and 245 DRPs during consultation (mainly ADEs, 58.4%). About three-fifths DRPs were caused by antibiotics. Comparing DRPs identified during pharmacy intervention and consultation, physicians consulted pharmacists more on questions related to medication safety, while pharmacists also paid attention to treatment effectiveness, which was consulted less commonly. Conclusion Implementation of PC model is beneficial in guiding pharmacy practice and improving efficacy especially under limited human resources. Physicians and pharmacists shall continue ensuring drug safety and be familiar with the scope of PC and clinical need for a better cooperation.