Skin-friction drag reduction in (water) turbulent boundary layers using bubble injection has been studied for some time. Ceccio (Annu Rev Fluid Mech 42:183–203, 2010) and Murai (Exp Fluids 55(7):1–28, 2014) have compiled drag reduction data from a number of different studies and facilities, and highlighted the large differences and scatter in the data even at the same bubble void fraction. Motivated by this, in the present work, we experimentally investigate within a single horizontal turbulent channel facility, drag modification using bubbles over a wide range of bubble void fraction (\(0< \alpha < 0.15\)), channel Reynolds number (22,500 \(< Re<\) 67,500), and the orientation of bubble injection (top/ bottom wall). In each of the cases, we have simultaneously measured drag modification and visualized the bubble dynamics. The drag modification is obtained from measurement of the mean pressure drop at four different vertical locations within the channel. The results show that even in the same facility, the drag reduction obtained at a fixed void fraction (\(\alpha\)) can be very different due to changes in bubble dynamics caused by changes in the other flow parameters. The visualizations show a number of bubble dynamics regimes depending on the parameters, with possibilities of both increased and decreased drag compared to the base (no bubble) case. The measurements for the bubble cases show significant vertical variations in the measured pressure drop within the channel, with these vertical variations being also dependent on the bubble distribution/dynamics. Interestingly, in some cases, the pressure drop at a given height even becomes negative, although the integrated pressure drop over the channel height, which is related to the overall drag, remains positive but lower than the base case. In terms of the overall drag, the top-wall injection is observed to give good drag reduction over a wide range of flow Re and \(\alpha\), but is seen to saturate beyond a threshold \(\alpha\). In contrast, the bottom-wall injection case shows that drag continuously decreases with \(\alpha\) at high channel Re, while at low channel Re, the drag is found to continually increase with \(\alpha\). The present study shows a maximum of about 60% increase and a similar 60% reduction in wall drag over the entire range of conditions investigated. For each of the bubble wall injection orientations (bottom/top/both wall), contour plots of drag modification and gain factor (fractional drag reduction per unit void fraction) are presented in the plane of \(\alpha\) and Re along with the corresponding bubble dynamics, which helps to delineate the different regimes seen in such bubbly channel flows.