This paper presents a study of signal crosstalk in a flip-chip quantum processor, demonstrating competitive crosstalk performance with other platforms. The processor consists of 25 superconducting qubits, with signal lines routed in a repeating pattern that is scalable to hundreds of qubits. The study focuses on two types of signal lines: capacitively coupled "xy-lines" for qubit drive and inductively coupled "z-lines" for flux control. The results show that the on-resonant crosstalk for xy-lines is better than -27 dB (average -37 dB), and the direct-current (dc) flux crosstalk for z-lines is less than 0.13% (average 0.05%). These levels are sufficiently small and show a decreasing trend with increasing distance, which is promising for further scaling. The study discusses the implications of these results for the design of low-crosstalk signal delivery architectures, including the influence of shielding structures, potential sources of crosstalk, and estimation of crosstalk-induced qubit-gate errors in scaled-up processors. The results indicate that the intrinsic crosstalk level due to direct capacitive interaction is better than the observed performance, and there is a trend of decreasing average crosstalk level with increasing qubit separation. The study also shows that the flux crosstalk is significantly reduced by using shielding structures, and that the crosstalk performance is competitive with other platforms. The paper concludes that careful routing and shielding of signal lines can already enable low crosstalk without active flux compensation, and that further investigations are needed to understand the source of crosstalk at the 0.1% level.This paper presents a study of signal crosstalk in a flip-chip quantum processor, demonstrating competitive crosstalk performance with other platforms. The processor consists of 25 superconducting qubits, with signal lines routed in a repeating pattern that is scalable to hundreds of qubits. The study focuses on two types of signal lines: capacitively coupled "xy-lines" for qubit drive and inductively coupled "z-lines" for flux control. The results show that the on-resonant crosstalk for xy-lines is better than -27 dB (average -37 dB), and the direct-current (dc) flux crosstalk for z-lines is less than 0.13% (average 0.05%). These levels are sufficiently small and show a decreasing trend with increasing distance, which is promising for further scaling. The study discusses the implications of these results for the design of low-crosstalk signal delivery architectures, including the influence of shielding structures, potential sources of crosstalk, and estimation of crosstalk-induced qubit-gate errors in scaled-up processors. The results indicate that the intrinsic crosstalk level due to direct capacitive interaction is better than the observed performance, and there is a trend of decreasing average crosstalk level with increasing qubit separation. The study also shows that the flux crosstalk is significantly reduced by using shielding structures, and that the crosstalk performance is competitive with other platforms. The paper concludes that careful routing and shielding of signal lines can already enable low crosstalk without active flux compensation, and that further investigations are needed to understand the source of crosstalk at the 0.1% level.