In this paper, we propose a detection algorithm to detect unknown partial-band noise interference (PBNI) which is frequently encountered in slow frequency-hopped spread spectrum (SFH-SS) systems. The algorithm utilizes a simple hypothesis testing to decide presence of PBNI in any particular hop after dehopping. A suitable threshold level is first derived to detect the PBNI by analyzing the statistical property of the outputs of two correlators used for quadrature phase-shift-keying (QPSK) demodulation. Then, the proposed detector is developed by summing the outputs of the correlators over one hop duration. We also formulate an expression for probability of detection (PD) and use it as a performance measure for detection. Analytical and simulation results show that the proposed detector is able to detect the PBNI very well even at low interference power (8 dB lower than signal power). In addition, numerical results reveal that the proposed algorithm can also be used to detect multitone interference (MTI) encountered in slow FH-SS systems.