Metabolic imaging using magnetic resonance spectroscopic imaging (MRSI) provides important biomarkers for brain neurochemistry. We developed a spiral-out-in (SOI) trajectory for human whole-brain MRSI at 7T to take advantage of increased sensitivity and spectral separation at ultra-high field. We hypothesized that spectral-spatial SOI sampling will provide higher signal-to-noise ratio(SNR) compared to spiral-out (SO) sampling by increasing the efficiency of data collection. We acquired data from phantom and six healthy volunteers. Metabolic maps, SNR, Cramér-Rao-Lower-Bounds (CRLB) were evaluated between SO and SOI acquisitions. By more efficient data points collection per repetition time, SOI provided a significant improvement in SNR and CRLB.