In the present work, we study the evolution of magnetic helicity in the solar corona. We compare the rate of change of a quantity related to the magnetic helicity in the corona to the flux of magnetic helicity through the photosphere and find that the two rates are similar. This gives observational evidence that helicity flux across the photosphere is indeed what drives helicity changes in the solar corona during emergence. For the purposes of estimating coronal helicity, we neither assume a strictly linear force-free field nor attempt to construct a nonlinear force-free field. For each coronal loop evident in extreme ultraviolet, we find a best-matching line of a linear force-free field and allow the twist parameter α to be different for each line. This method was introduced and its applicability discussed in Malanushenko et al. The object of this study is emerging and rapidly rotating AR 9004 over about 80 hr. As a proxy for coronal helicity, we use the quantity α i L i /2 averaged over many reconstructed lines of magnetic field. We argue that it is approximately proportional to the "flux-normalized" helicity H/Φ 2 , where H is the helicity and Φ is the total enclosed magnetic flux of the active region. The time rate of change of such a quantity in the corona is found to be about 0.021 rad hr −1 , which is comparable with the estimates for the same region obtained using other methods, which estimated the flux of normalized helicity to be about 0.016 rad hr −1 .