Studies of the primary care treatment of depressed elderly patients are con
strained by limited time and space and by subject burden. Research assessme
nts must balance these constraints with the need for obtaining clinically m
eaningful information. Due to the wide-ranging impact of depression, assess
ments should also focus on suicidality, hopelessness, substance abuse, anxi
ety, cognitive functioning, medical comorbidity, functional disability, soc
ial support, personality, service use and satisfaction with services. This
paper describes considerations concerning the assessment selection process
fur primary care studies, using the PROSPECT (Prevention of Suicide in Prim
ary Care Elderly: Collaborative Trial) study as an example. Strategies are
discussed for ensuring that data are complete, valid and reliable. Copyrigh
t (C) 2001 John Wiley & Sons, Ltd.