Ever since the first attempts to fit Jicamarca autocorrelation function (ACF) measurements in the 1970s using a full nonlinear least squares analysis, an apparent electron-ion temperature ratio below unity has been deduced for a large portion of the F region data. The cause of this unexpected and geophysically unreasonable result has been a mystery until recently, when Sulzer and Gonz¿lez [1999> (herein SG) explained how electron Coulomb collisions can distort, or narrow, the incoherent backscatter spectrum, and that for this narrowing to be observable two conditions must be met. First, the radar k vector must lie in a small range near perpendicular to the magnetic field, and second, the radar wavelength must be sufficiently long. Both of these conditions are true at Jicamarca. The accurate calculations from the SG theory are now available in a compact library, which we have incorporated into an incoherent scatter least squares fitting code. Using this code, we have reduced Jicamarca ACF data taken with the Faraday double-pulse mode, and find that the SG theory correctly interprets the ACF data from Jicamarca, thereby solving the longstanding Te/Ti ratio problem and thus allowing accurate electron and ion temperature measurements. ¿ 2001 American Geophysical Union