This controlled study compared the strength and porosity of 48 polymethylmethacrylate cement-implant constructs prepared with open bowl versus vacuum mix technique. Forty-eight blast finished stainless steel rods of 13 mm diameter were implanted with centralizers into 17-mm inner diameter tubes that had been retrograde filled with polymethylmethacrylate cement. The eight cement preparations used were open bowl and vacuum mixed Simplex, Osteobond, Zimmer Dough Type, or Palacos R. Six replications of each condition were performed. The tubes were maintained at 37 degrees C. Each tube was cut transversely into five segments. The center three segments were used for data analysis: pushout strength, cycles to failure, and interface porosity analysis. Rod pushout data showed there was no significant difference between open bowl and vacuum mixed samples when all cement brands were combined. Mean sheer force for Palacos R vacuum mixed samples was greater than open bowl (634+/-47 versus 423+/-171), whereas the force for the Zimmer Dough Type cement open bowl was greater than that of the vacuum mixed samples (901+/-71 versus 705 +/-82). Cycles to failure data did not show significant differences when open bowl and vacuum mixed samples were compared when cements were analyzed individually or combined. Image analysis of cement-implant interfaces showed that vacuum mixing reduced void area significantly compared with open bowl mixing in the Palacos R and Osteobond preparations. Vacuum mixing does not appear to reduce cement prosthesis interface porosity or improve its mechanical properties in all cements.