The objective of this study was to compare the shear-peel band strength of 5 orthodontic cements using both factory and in-office micro-etched bands. The 5 orthodontic cements evaluated were a zinc phosphate (Fleck's Cement), 2 resin-modified glass ionomer cements (RMGI)(3M Multicure glass ionomer and Optiband), and 2 polyacid-modified composite resin cements (PMCR)(Transbond Plus and Ultra Band Lok). Salivary contamination was examined with a polyacid-modified composite resin (Transbond Plus). Two hundred and eighty extracted human molar teeth were embedded in resin blocks and each was randomly assigned to the following 7 groups: 6 groups with factory etched bands, 5 cement groups and salivary contaminated group, and 1 in-office micro-etched group. The cemented teeth were put in deionized water at 37 degrees C for 30 days and thermocycled for 24 hours. The force required to break the cement bond was used as a measure of shear-peel band retention. With the use of an Instron testing machine, a shear-peel load was applied to each cemented band. Data were analyzed with a one-way analysis of variance (ANOVA) with a Tukey test for the multiple comparisons. The RMGIs and PMCRs demonstrated significantly greater shear-peel band strengths compared to the zinc phosphate cement. No statistically significant differences were noted between the RMGI cement and PMCR cements and within the RMGI groups, however, there was a statistically significant difference within the PMCR groups. Significantly lower band strengths were noted with the saliva contaminated PMCR cement group (Transbond Plus) and the inpractice sandblasted PMCR group. Both RMGIs and PMCRs were found to demonstrate favorable banding qualities. The lower band strength with saliva-contaminated bands suggests that moisture control is critical when using a PMCR. The variability noted in the in-office micro-etched bands might be technique related.