Timely linkage to HIV medical care has the potential to improve individual health outcomes and prevent secondary HIV transmission. Recent research found that estimates of delayed care entry varied by study design, with higher estimates among studies using only HIV case surveillance data. In this analysis, we compared the prevalence and risk factors for care delay using data from two studies with different designs conducted in New York City. The Medical Monitoring Project (MMP) used a retrospective design to estimate historical delay among persons currently receiving care, while the Never in Care (NIC) study used a prospective design to estimate current delay status among persons who were care-naive at baseline. Of 513 MMP subjects in 2007-2008, 23% had delayed care entry greater than three months after diagnosis. Independent risk factors for care delay were earlier year of diagnosis and testing positive in a nonmedical environment. Of 28 NIC subjects in 2008-2010, over half had tested positive in a nonmedical environment. The primary-stated reasons for delay were the same in both studies: denial of HIV status and lack of perceived need for medical care. The strengths and weaknesses of surveillance only, prospective, and retrospective study designs with respect to investigating this issue are explored. Future studies and interventions should be mindful of the common selection biases and measurement limitations with each design. A triangulation of estimates from varying designs is suggested for accurately measuring care linkage efforts over time.