Background: In sub-Saharan Africa, there are no validated screening tools for delirium in older adults. This study assesses clinical utility of two instruments, the IDEA cognitive screen and the Confusion Assessment Method (CAM) for identification of delirium in older adults admitted to medical wards of a tertiary referral hospital in Tanzania. Method: The IDEA cognitive screen and CAM were administered to a consecutive cohort of older individuals on admission to Kilimanjaro Christian Medical Centre using a blinded protocol. Consensus diagnosis for delirium was established against DSM-V criteria and dementia by DSM-IV criteria Results: Of 507 admission assessments, 95 (18.7%) had DSM-V delirium and 95 (18.7%) had DSM-IV dementia (33 (6.5%) delirium superimposed on dementia). The CAM and IDEA cognitive screen had very good diagnostic accuracy for delirium (AUROC curve 0.94 and 0.87 respectively). However, a number of participants (10.5% and 16.4% respectively) were unable to complete these screening assessments due to reduced consciousness, or other causes of reduced verbal response and were excluded from this analysis; many of whom met DSM-V criteria for delirium. Secondary analysis suggests that selected cognitive and observational items from the CAM and IDEA cognitive screen may be as effective as the full screening tools in identifying delirium even in unresponsive patients. Conclusion: Both instruments appeared useful for delirium screening in this inpatient setting, but had significant limitations. The combination of assessment items identified may form the basis of a brief, simple delirium screening tool suitable for use by non-specialist clinicians. Further development work is needed.