In vivo calcification of polyethylene glycol diacrylate (PEG DA) hydrogels of molecular weight (MW) 400, 1000, 4000, 6000 and 10,000 and polyethylene glycol tetraacrylate (PEG TA) of MW 18,500 was investigated using a rat subcutaneous model. This study was performed in 4-wk-old rats for durations of 1, 3, 6 and 8 wk. The results indicate a strong dependence of calcification upon the MW of the PEG precursor or the MW between crosslinks. Results for gels implanted for 6 wk show that calcification was maximal at a PEG MW of 1000 (224 mg/g +/- 12.8, n = 4) (mean +/- SEM) with less at MW = 400 (23.0 mg/g +/- 9.30, n = 4) and considerably less at higher MWs, e.g. for MW = 10,000 (0.23 mg/g +/- 0.01, n = 4). Results for other time periods indicate a similar calcification trend. The extent of calcification of gels from PEG TA (MW = 18,500) was intermediate (1.09 mg/g +/- 0.43, n = 3) between PEG DA (MW = 6000) (1.39 mg/g +/- 0.42, n = 6) and PEG DA (MW = 10,000) at 6 wk, i.e. calcification depended upon the PEG MW between crosslinks. When composite gels were implanted, such that a highly calcifying gel (MW = 400 or 1000) was encapsulated within a gel of low calcification (MW = 4000), the gel inside calcified to at least the same extent as if it had not been encapsulated. Thus, direct contact with tissues is apparently not necessary for calcification to occur. Energy dispersive X-ray spectroscopy was performed on the mineral deposits in the gels and a P:Ca ratio of 0.67 +/- 0.04 (95% confidence interval) for MW 1000 gels and 0.60 +/- 0.07 for MW 400 gels was found to be consistent with deposition of Ca3(PO4)2.