P1dB and two-tone IMD are separate nonlinear tests. P1dB uses one tone and a power sweep. It compares measured gain with a stated small-signal reference and reports the input- or output-referred level where gain has fallen by 1 dB. The plan must define frequency, sweep direction, dwell, bias, thermal state, load, reference planes and how the linear gain fit is established.
A two-tone test applies two frequencies with stated spacing and per-tone power, then measures fundamentals and intermodulation products, commonly the third-order products near the tones. State whether power is per tone or combined, provide source isolation, and show that source IMD and receiver distortion are below the DUT products. A residual-system measurement with a through path or suitable attenuation is part of the evidence.
Do not use an extrapolated IP3 as a substitute for compression, or either test as a substitute for modulated EVM, adjacent-channel power or memory-effect testing. Compare devices only when reference planes, tone spacing, power convention, bandwidth, averaging, bias and temperature match. If the real application uses pulsed or modulated signals, add that waveform rather than assuming the two-tone result predicts it.