Purpose: CD4 count plays a central role in CDC staging for HIV/AIDS. Guidelines based on CD4 counts, however, can't be used in resource-limited settings where CD4 counts are impossible or too expensive to obtain. For example, in a clinical cohort from Chennai, India recruited from 1996 to 2004, 41% of adult HIV positive patients did not have a single CD4 count. In this type of situations, physicians face questions such as when to start and when to stop treatment based on clinical evidence alone. Opportunistic infections (OI) are clinic outcomes accompanying advanced HIV/AIDS. It has been shown that different OIs occur at different CD4 counts levels. In the US, the incidence of OIs has shown a striking decline in recent years due to the availability of new drugs and improved HIV/AIDS care. However in many developing countries, OIs are still widespread among HIV/AIDS patients. In resource-limited clinical settings, patients often present for care when they already have some type of OIs. In this study we give a quantified approach to assess CD4 counts for HIV infected patients based on patient's history of OIs. We show that different types of OIs and, for patients with more than one type of OIs, different patterns of OIs correspond to different CD4 count levels. Data: We base our estimation on data from the Multicenter AIDS Cohort Study (MACS).We focus our study on adult (age 18 or above), HIV infected and antiretroviral treatment naïve patients. Our analysis included 1359 patients' records. Methods: We empirically estimate mean and median CD4 count for patiens with history of different type of OIs. We also estimate distributions of CD4 counts for patients with two different types of OIs. We give 95% confidence intervals of each estimate. Results: We found that history of MAC and CMV show the strongest evidence of severe disease by indicating the lowest CD4 count level (<70cells/l). TB, Kaposi, fungal, toxo, and PCP all predict CD4 counts under 200cells/l. Patients with a history of two different types of OIs tend to have CD4 count below 50cells/l. Conclusion: When CD4 count is unavailable, different type and patterns of opportunistic infections provide valuable information about patients' CD4 level.