Real ear SPL (RE/SSPL90) produced by four different hearing aids was compared to the standard SSPL90 measurements for the same hearing aids on a 2-cc coupler. The effects of venting, location, and reliability for RE/SSPL90 were also studied. As expected, the real ear output level was greater than the coupler level. No systematic influence of earmold venting was found. An earmold that occludes an ear canal decreases the output in the high-frequency range so that only a small difference remains between RE/SSPL90 and 2-cc coupler levels above 2000 Hz. The difference seems to be affected slightly by the output impedance of the hearing aid receiver system. Individual variation for the RE/SSPL90 is as much as 10 dB. Finally, the saturated output levels measured were found to be insensitive to changes in actual microphone input levels caused by different procedures for calibrating ambient SPL. The RE/SSPL90 method is easy to use and reliability seems quite adequate for the typical clinical setting.