Active Constellation Extension (ACE) is widely used to reduce Peak-to-Average Power Ratio (PAPR) in Offset Quadrature Amplitude Modulation based Orthogonal Frequency Division Multiplexing (OFDM/OQAM). To improve overall performance and energy efficiency in OFDM/OQAM systems, this paper proposes Adaptive Scaling (AS) for effective ACE (AS-ACE). The proposed scheme exploits the overlapping structure of the OFDM/OQAM signals and jointly considers adjacent data blocks to obtain the clipping noise. The proposed scheme adaptively extends the constellation points to effectively eliminate peaks in OFDM/OQAM signals. This is achieved by using peak samples of clipping noise, used to generate the peak correcting signal. Least square approximation algorithm is used to fit the waveform of peak correcting signal to the waveform of clipping noise without any decrease in the constellation distance. Simulation results show that the AS-ACE reduces PAPR and Bit Error Rate (BER) and improves convergence speed compared to OSGP technique without requiring increased processing or additional information at the receiver end.