UGC3995 is an interacting and occulting galaxy pair. UGC3995B is a foreground face-on spiral and UGC 3995A a bright background spiral with an AGN. We present analysis of the dust in the disc of UGC 3995B based on archival Hubble Space Telescope (HST) WFPC2 and PPAK IFU data from the CALIFA survey's first data release. From the HST F606W image, we construct an extinction map by modeling the isophotes of the background galaxy UGC 3995A and the resulting transmission through UGC 3995B. This extinction map of UGC 3995B shows several distinct spiral extinction features. The radial distribution of AV values declines slowly with peaks corresponding to the spiral structures. The distribution of A(V) values in the HST extinction map peaks near A(V) = 0.3-0.4. Beyond this point, the distribution of A(V) values drops like an exponential: N(A(V)) = N(0)xe((-AV/0.5)). The 0.5 value is higher than typical for a spiral galaxy. The outer arms may be tidally distended; the extinction in the corresponding interarm regions is small to an unusually small radius. To analyze the PPAK IFU data, we take the ratio of a fibre spectrum in the overlap region and the corresponding background fiber spectrum to construct an extinction curve. We fit the Cardelli, Clayton and Mathis (CCM) curve to the extinction curve of each fiber element in the overlap region. A map of the extinction constructed from PPEX IFU data-cubes shows the same spiral structure of the HST extinction map but the some differences in the distribution of the normalization of the CCM fits (A(V)). The inferred extinction slopes (R-V) maps do not display any structure and a range of values partly due to the sampling effects of the disc by fibers, sometimes due to bad fits, and possibly partly due to some reprocessing of dust grains in the interacting disc. We compare these findings to our other analysis of an occulting pair with HST and IFU data. In both cases the canonical R-V = 3.1 is not recovered even though there is enough signal in the extinction curve. We attribute this to mixing opaque and more transparent sections of the disc in each resolution element (similar to 3 '' or 0.9 kpc). To illustrate the difficulty of imposing a R-V = 3.1 law over a section of a spiral disc, we average all spectra and show how a fully gray extinction curve is recovered.