BACKGROUND: Tuberculosis (TB) presents a serious problem in Mozambique. HIV prevalence among TB patients is estimated at 47%. A delay in having their first CD4+ cell count could lead to a missed opportunity for ART initiation due to a CD4+ cell increase above the cut-off caused by TB treatment. The objective is to describe CD4+ cell response during TB treatment and quantify the effect of TB treatment and ART on this response. METHODS: All new HIV + adult TB cases in 2007 from three TB clinics in Mozambique were included. Data on TB diagnosis and treatment and HIV parameters were collected. A general mixed model was used for CD4+ cell count response. RESULTS: 338 HIV + patients were notified and 252 (75%) were included in the analysis. Using TB medication was not independently associated with the CD4+ count response (19 cells/mm3; 95% CI: -40 to 79; p = 0.529). ART-use was associated with statistically significantly higher CD4+ cells compared to no ART-use (81 cells/mm3; 95% confidence interval (CI): 12 to 151; p = 0.022). CONCLUSION: In this study, no independent effect of TB treatment on CD4+ cell count was found. HIV-infected TB patients on ART had a significantly higher CD4+ cell count than those not receiving ART. CD4+ cell counts for patients not on ART at TB treatment start, remained below the cut off for initiating ART during the first three months of TB treatment; therefore some delay in getting the first CD4+ cell count would not lead to missing the opportunity to start ART.