Two series of 100 consecutive primary total hip arthroplasties, each using a single design of noncemented or cemented femoral component (all 28 mm heads), were compared. One cemented and two noncemented stems underwent revision for aseptic loosening. Of unrevised hips, outcome data statistically favored cemented, rather than noncemented, stems. The data for cemented and noncemented stems, respectively, were: An excellent to good result in 97% versus 88%; thigh pain in 3% versus 40%; subsidence in 0% versus 22%; and endosteal cavitation in 6% versus 12%. For patients with 25 unrevised matched pairs, selected by gender, age, diagnosis, and weight, outcome data also statistically favored cemented over noncemented stems, respectively: an excellent or good result in 25 versus 20 hips; thigh pain in two versus eight hips; and subsidence in none versus six hips. Midterm followup data for these concurrent total hip arthroplasty series of a mid 1980s design revealed prevalence of mechanical failure of 1% for cemented stems and 4% for noncemented stems. Corroborating matched pair comparison neutralized selection bias as a causative factor for these differences. These data indicate contemporary cemented femoral stem fixation is superior to second generation noncemented femoral stem fixation. Controlled comparative studies at midterm to long term followup, such as in this report, are needed to define outcome and indications for current third generation noncemented stem fixation.