A curved area with embossment on an analgesic tablet with three active pharmaceutical ingredients (APIs) was imaged with a Raman instrument equipped with LiveTrack™ and StreamLine™ options that allow for continuous focus adjustment and fast acquisitions of micro-Raman spectra. The mapping spectra were acquired in regular- and high-confocal mode with varying time acquisitions in high-confocal maps. Univariate imaging, self-modelling curve resolution (SMCR), and principal component analysis (PCA) were used to produce images of all three APIs that were then combined in composite images. Univariate images were found to be most sensitive to signal to noise ratio in the spectra. The images derived by using SMCR appear to be of highest quality and most reliable but the lowest concentration API (caffeine) was not successfully resolved by SMCR. PCA-derived images were based on ambiguous loadings that are combinations of various bands of different sign. Caffeine distribution was still correctly displayed despite its bands not being major contributors in any of the loadings. Interestingly, due to its low concentration vs. the three APIs in the tablet, avicel was very difficult to identify and image, and only tiny spectral responses with a few correlated random pixels of avicel were identified in PC score images.