Histological analysis is meaningful in diagnosis only if the targeted tissue is obtained in the biopsy. Often, physicians have to take a tissue sample without accurate information about the location of the instrument tip. A novel biopsy needle with bioimpedance-based tissue identification has been developed to provide data for the automatic classification of the tissue type at the tip of the needle. The aim of this study was to examine the resolution of this identification method and to assess how tissue heterogeneities affect the measurement and tissue classification. Finite element method simulations of bioimpedance measurements were performed using a 3D model. In vivo data of a porcine model were gathered with a moving needle from fat, muscle, blood, liver, and spleen, and a tissue classifier was created and tested based on the gathered data. Simulations showed that very small targets were detectable, and targets of 2 × 2 × 2 mm3 and larger were correctly measurable. Based on the in vivo data, the performance of the tissue classifier was high. The total accuracy of classifying different tissues was approximately 94%. Our results indicate that local bioimpedance-based tissue classification is feasible in vivo, and thus the method provides high potential to improve clinical biopsy procedures.