This paper proposes a power allocation method for massive MIMO-ISAC systems to achieve dual-function transmit beamforming. The authors propose a structured transmit beamformer that is a weighted combination of a linear precoder (maximum ratio or zero-forcing) and a pre-designed array beamformer for sensing. The weights are determined by the power allocated to the beamformers, and the design problem is formulated as a total transmit power minimization problem that satisfies the requirements for both communication and sensing. The use-and-then-forget strategy is leveraged to derive simplified performance metrics for communication and sensing, leading to linear programming (LP) problems for the design. Analytical solutions are derived, making the design computationally attractive. Simulations demonstrate the effectiveness and performance of the proposed methods. The main contributions include the proposal of a structured mMIMO-ISAC transmit beamformer, derivation of closed-form expressions for effective SINR and mainlobe-to-average-sidelobe ratio (MASR), formulation of LP problems for power allocation, and a detailed performance analysis. The results show that the proposed design is computationally efficient and can achieve satisfactory sensing and communication performance. The minimum number of antennas required to meet specified SINR and MASR requirements is also determined. The analysis is extended to correlated channels, showing that the proposed method is robust to channel correlation. The results indicate that the proposed power allocation method is effective for achieving dual-function transmit beamforming in mMIMO-ISAC systems.This paper proposes a power allocation method for massive MIMO-ISAC systems to achieve dual-function transmit beamforming. The authors propose a structured transmit beamformer that is a weighted combination of a linear precoder (maximum ratio or zero-forcing) and a pre-designed array beamformer for sensing. The weights are determined by the power allocated to the beamformers, and the design problem is formulated as a total transmit power minimization problem that satisfies the requirements for both communication and sensing. The use-and-then-forget strategy is leveraged to derive simplified performance metrics for communication and sensing, leading to linear programming (LP) problems for the design. Analytical solutions are derived, making the design computationally attractive. Simulations demonstrate the effectiveness and performance of the proposed methods. The main contributions include the proposal of a structured mMIMO-ISAC transmit beamformer, derivation of closed-form expressions for effective SINR and mainlobe-to-average-sidelobe ratio (MASR), formulation of LP problems for power allocation, and a detailed performance analysis. The results show that the proposed design is computationally efficient and can achieve satisfactory sensing and communication performance. The minimum number of antennas required to meet specified SINR and MASR requirements is also determined. The analysis is extended to correlated channels, showing that the proposed method is robust to channel correlation. The results indicate that the proposed power allocation method is effective for achieving dual-function transmit beamforming in mMIMO-ISAC systems.