Magma-rock interactions involve open-system processes where rocks undergo partial melting, minerals dissolve into the melt, and magmas hybridize. In this context, it is poorly understood how felsic rocks characterized by low-δ18O, similar to hydrothermally altered crust, form through reactions with basaltic magma or by the differentiation of hybridized magma. Our aim was to understand the consequences of low-δ18O-felsic crust interacting with basaltic melt in shallow crustal settings and to explain the origin of low-δ18O rhyolite and associated glasses encountered within the active Krafla volcano (KJ-39 and IDDP-1), Northern Iceland (Figure 1). Here, we report the results of an experimental study on the rate and mechanism of the reaction between anhydrous Fe-Ti tholeiite basaltic melt and felsic granophyre rock at 1250°C and 0.2 GPa, water activity, aH2O = 0 and with oxygen fugacity fixed by the QFM (quartz-fayalite-magnetite) mineral buffer. We applied SEM, HR-EPMA and HR-SIMS as the micro-analytical methods to investigate the experimental and natural phases. Our experiments (Table 1) reveal that the granophyre rock partially melts and dissolve (Figure 2), with the dissolution controlled by Si diffusion into basaltic melt. Comparing the major element and oxygen isotope chemistry of the Krafla glasses with the partial and interface melts produced in our experiments shows similarity (Figures 3-4, Table 2). Our data suggest that zero-age rhyolites are likely produced through partial melting of quartz-feldspar-rich granophyre rocks, accompanied by hybridization driven by silica diffusion in the basaltic magma (Figure 5). If the dissolution of felsic rock in basaltic magma reservoir were uniquely diffusion-controlled and restricted to interaction along the interface, very slow hybridization of the interface melt with basaltic melt would occur (Figure 6). We propose that to produce the natural glasses observed in boreholes KJ-39 and IDDP-1 of Krafla volcano, the parental granophyre rock was progressively melted due to interaction with a hot basaltic magma (Figure 7), allowing the formation of chemically heterogeneous magma, which was sampled in 2008 and 2009 on the same depths by drilling (Figures 3-4, Table 2).