Purpose: Past research has established parameters for the 2f1-f2 distortion product otoacoustic emissions (DPOAEs) that enhance response levels (e.g., L1 - L2 = 10 dB; f2/f1 = 1.22; L1, L2 = 65, 55 dB SPL). These same parameters do not optimize 2f2-f1 DPOAEs. Therefore, this study was conducted to evaluate more completely those parameters that produce the most robust 2f2-f1 output. Method: Input-output functions of the 2f2-f1 component were obtained from 20 normal-hearing adults (with f2 = 2000 Hz and 4000 Hz). Frequency ratios, level differences, and overall sound levels were manipulated with parameter combinations chosen to expand on information established in previous studies. The DPOAE signal-to-noise ratio (SNR), response presence, and output levels were measured. Results: In general, mean SNRs and 2f2-f1 levels were greater, and DPOAEs were present more often for 2000 than for 4000 Hz across all parameter combinations. No single parameter combination resulted in pronounced maxima for 2f2-f1, which is consistent with past studies but is in sharp contrast to results for the more familiar 2f1-f2. Conclusions: Overall, the results suggest that to maximize the 2f2-f1 level, lower test frequencies, low frequency ratio, essentially equal levels of primary tones, and moderate stimulus levels are best, and such parameters might be important for applying DPOAE measurement to assessment of normal and impaired auditory function.