Abstract: To address the challenges of large array aperture and high array element redundancy in traditional nested vertical linear arrays for radiated noise measurement, a broadband radiated noise measurement method for underwater targets based on a sparse nested vertical linear array is proposed. Firstly, based on the array element reuse concept of traditional nested arrays, a nested array-integrated constant beamwidth beamforming strategy building upon convex optimization algorithms is adopted to optimize the design of a 2-octave sparse nested array structure, and the optimized beamforming weighting vector is obtained. Then, the weighting vector is modified by using the direct wave channel transfer function between the target’s equivalent sound center and the nested array, so that the constant beamwidth beamforming method can be adapted to the near-field measurement of target radiated noise, and the measurement accuracy of underwater target radiated noise spectrum source level and sound source level is improved. Subsequently, the weighting vector is modified by using the direct wave channel transfer function between the target’s equivalent acoustic center and the nested array, so that the constant beamwidth beamforming method can be adapted to the near-field measurement of target radiated noise, and the measurement accuracy of underwater target radiation noise level. Both simulation calculations and lake test processing results show that compared with the traditional nested array, the sparse nested array significantly reduces the array aperture and has fewer array elements while ensuring the measurement performance. In addition, its uncertainty of radiated noise measurement is about 1.9 dB.